README.md

metadata

dataset_info:
  - config_name: default
    features:
      - name: utterance
        dtype: string
      - name: label
        dtype: int64
    splits:
      - name: train
        num_bytes: 715028
        num_examples: 10003
      - name: test
        num_bytes: 204010
        num_examples: 3080
    download_size: 378619
    dataset_size: 919038
  - config_name: intents
    features:
      - name: id
        dtype: int64
      - name: name
        dtype: string
      - name: tags
        sequence: 'null'
      - name: regexp_full_match
        sequence: 'null'
      - name: regexp_partial_match
        sequence: 'null'
      - name: description
        dtype: 'null'
    splits:
      - name: intents
        num_bytes: 3420
        num_examples: 77
    download_size: 4651
    dataset_size: 3420
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: test
        path: data/test-*
  - config_name: intents
    data_files:
      - split: intents
        path: intents/intents-*

banking77

This is a text classification dataset. It is intended for machine learning research and experimentation.

This dataset is obtained via formatting another publicly available data to be compatible with our AutoIntent Library.

Usage

It is intended to be used with our AutoIntent Library:

from autointent import Dataset

banking77 = Dataset.from_datasets("AutoIntent/banking77")

Source

This dataset is taken from PolyAI/banking77 and formatted with our AutoIntent Library:

"""Convert events dataset to autointent internal format and scheme."""

import json

import requests
from datasets import Dataset as HFDataset
from datasets import load_dataset

from autointent import Dataset
from autointent.schemas import Intent, Sample


def get_intents_data(github_file: str | None = None) -> list[Intent]:
    """Load specific json from HF repo."""
    github_file = github_file or "https://huggingface.co/datasets/PolyAI/banking77/resolve/main/dataset_infos.json"
    raw_text = requests.get(github_file, timeout=5).text
    dataset_description = json.loads(raw_text)
    intent_names = dataset_description["default"]["features"]["label"]["names"]
    return [Intent(id=i, name=name) for i, name in enumerate(intent_names)]


def convert_banking77(
    banking77_split: HFDataset, intents_data: list[Intent], shots_per_intent: int | None = None
) -> list[Sample]:
    """Convert one split into desired format."""
    all_labels = sorted(banking77_split.unique("label"))

    n_classes = len(intents_data)
    if all_labels != list(range(n_classes)):
        msg = "Something's wrong"
        raise ValueError(msg)

    classwise_samples = [[] for _ in range(n_classes)]

    for sample in banking77_split:
        target_list = classwise_samples[sample["label"]]
        if shots_per_intent is not None and len(target_list) >= shots_per_intent:
            continue
        target_list.append(Sample(utterance=sample["text"], label=sample["label"]))

    samples = [sample for samples_from_one_class in classwise_samples for sample in samples_from_one_class]
    print(f"{len(samples)=}")
    return samples


if __name__ == "__main__":
    intents_data = get_intents_data()
    banking77 = load_dataset("PolyAI/banking77", trust_remote_code=True)

    train_samples = convert_banking77(banking77["train"], intents_data=intents_data)
    test_samples = convert_banking77(banking77["test"], intents_data=intents_data)

    banking77_converted = Dataset.from_dict(
        {"train": train_samples, "test": test_samples, "intents": intents_data}
    )