Roberta With Kornli

🚀 Zero-shot Classification Model for Korean NLI

This model is designed for zero-shot classification in the Korean language. It fine-tunes the klue/roberta-base model on the mnli and xnli datasets of kor_nli, offering high accuracy in text classification tasks.

🚀 Quick Start

Prerequisites

This model has been referred to the following link: https://github.com/Huffon/klue-transformers-tutorial.git

Model Details

Training Parameters

Code Modification for Zero-shot Pipeline

For models that do not use token_type_ids, such as RoBERTa, the zero-shot pipeline cannot be directly applied (as of transformers==4.7.0). Therefore, you need to add the following code for conversion. This code is also a modification of the code from the above GitHub repository.

class ArgumentHandler(ABC):
    """
    Base interface for handling arguments for each :class:`~transformers.pipelines.Pipeline`.
    """

    @abstractmethod
    def __call__(self, *args, **kwargs):
        raise NotImplementedError()


class CustomZeroShotClassificationArgumentHandler(ArgumentHandler):
    """
    Handles arguments for zero-shot for text classification by turning each possible label into an NLI
    premise/hypothesis pair.
    """

    def _parse_labels(self, labels):
        if isinstance(labels, str):
            labels = [label.strip() for label in labels.split(",")]
        return labels

    def __call__(self, sequences, labels, hypothesis_template):
        if len(labels) == 0 or len(sequences) == 0:
            raise ValueError("You must include at least one label and at least one sequence.")
        if hypothesis_template.format(labels[0]) == hypothesis_template:
            raise ValueError(
                (
                    'The provided hypothesis_template "{}" was not able to be formatted with the target labels. '
                    "Make sure the passed template includes formatting syntax such as {{}} where the label should go."
                ).format(hypothesis_template)
            )

        if isinstance(sequences, str):
            sequences = [sequences]
        labels = self._parse_labels(labels)

        sequence_pairs = []
        for label in labels:
            # Modified part: To prevent the automatic attachment of `token_type_ids` when two sentences are input as a pair,
            # connect the two sentences in advance based on the `sep_token`.
            sequence_pairs.append(f"{sequences} {tokenizer.sep_token} {hypothesis_template.format(label)}")

        return sequence_pairs, sequences

Applying the Modified Code

You need to apply the above code when defining the classifier.

classifier = pipeline(
    "zero-shot-classification",
    args_parser=CustomZeroShotClassificationArgumentHandler(),
    model="pongjin/roberta_with_kornli"
)

💻 Usage Examples

Basic Usage

sequence = "배당락 D-1 코스피, 2330선 상승세...외인·기관 사자"	
candidate_labels =["외환",'환율', "경제", "금융", "부동산","주식"]

classifier(
    sequence,
    candidate_labels,
    hypothesis_template='이는 {}에 관한 것이다.',
)

>>{'sequence': '배당락 D-1 코스피, 2330선 상승세...외인·기관 사자',
 'labels': ['주식', '금융', '경제', '외환', '환율', '부동산'],
 'scores': [0.5052872896194458,
  0.17972524464130402,
  0.13852974772453308,
  0.09460823982954025,
  0.042949128895998,
  0.038900360465049744]}

📄 License

This model is licensed under the Apache-2.0 license.