Summary

Summary

The Summary pipeline summarizes text. This pipeline runs a text2text model that abstractively creates a summary of the input text.

Example

The following shows a simple example using this pipeline.

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

See the link below for a more detailed example.

Notebook	Description
Building abstractive text summaries	Run abstractive text summarization

Configuration-driven example

Pipelines are run with Python or configuration. Pipelines can be instantiated in configuration using the lower case name of the pipeline. Configuration-driven pipelines are run with workflows or the API.

config.yml

# Create pipeline using lower case class name
summary:
# Run pipeline with workflow
workflow:
  summary:
    tasks:
      - action: summary

Run with Workflows

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

Run with API

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"]}'

Methods

Python documentation for the pipeline.

Source code in txtai/pipeline/text/summary.py

15
16

def init(self, path=None, quantize=False, gpu=True, model=None, kwargs):
    super().init(“summarization”, path, quantize, gpu, model, kwargs)

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.

Parameters:

Name	Description	Default
`text`	text\|list	required
`minlength`	minimum length for summary	`None`
`maxlength`	maximum length for summary	`None`
`workers`	number of concurrent workers to use for processing data, defaults to None	`0`

Returns:

Type	Description
	summary text

Source code in 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