autotrain-data-processor
Processed data from AutoTrain data processor ([2023-09-07 18:33 ]
c0d0625
metadata
language:
  - en
task_categories:
  - summarization

AutoTrain Dataset for project: yetipy

Dataset Description

This dataset has been automatically processed by AutoTrain for project yetipy.

Languages

The BCP-47 code for the dataset's language is en.

Dataset Structure

Data Instances

A sample from this dataset looks as follows:

[
  {
    "feat_code": "        \n        return self.correctwords(originalwords, [newword], **kwargs)",
    "target": "def mergewords(self, newword, *originalwords, **kwargs)",
    "text": "TODO: Write documentation",
    "feat_loss_without_docstring": 9.7212610245,
    "feat_loss_with_docstring": 9.146777153,
    "feat_factor": 1.0628072448
  },
  {
    "feat_code": "        \n\n        if size == 0: return [] #for efficiency\n\n        context = []\n        e = self\n        while len(context) < size:\n            e = e.previous(True,scope)\n            if not e: break\n            context.append(e)\n\n        if placeholder:\n            while len(context) < size:\n                context.append(placeholder)\n\n        context.reverse()\n        return context",
    "target": "def leftcontext(self, size, placeholder=None, scope=None)",
    "text": "Returns the left context for an element, as a list. This method crosses sentence/paragraph boundaries by default, which can be restricted by setting scope",
    "feat_loss_without_docstring": 3.7681286335,
    "feat_loss_with_docstring": 3.9110825062,
    "feat_factor": 0.9634490266
  }
]

Dataset Fields

The dataset has the following fields (also called "features"):

{
  "feat_code": "Value(dtype='string', id=None)",
  "target": "Value(dtype='string', id=None)",
  "text": "Value(dtype='string', id=None)",
  "feat_loss_without_docstring": "Value(dtype='float64', id=None)",
  "feat_loss_with_docstring": "Value(dtype='float64', id=None)",
  "feat_factor": "Value(dtype='float64', id=None)"
}

Dataset Splits

This dataset is split into a train and validation split. The split sizes are as follow:

Split name Num samples
train 80
valid 20