Class SST2Processor

src/process_data.py:204–231

`SST-2 `_ dataset is a dataset for sentiment analysis. It is a modified version containing only binary labels (negative or somewhat negative vs somewhat positive or positive with neutral sentences discarded) on top of the original 5-labeled dataset

class SST2Processor(DataProcessor):
    """
    `SST-2 <https://nlp.stanford.edu/sentiment/index.html>`_ dataset is a dataset for sentiment analysis. It is a modified version containing only binary labels (negative or somewhat negative vs somewhat positive or positive with neutral sentences discarded) on top of the original 5-labeled dataset released first in `Recursive Deep Models for Semantic Compositionality Over a Sentiment Treebank <https://aclanthology.org/D13-1170.pdf>`_

    We use the data released in `Making Pre-trained Language Models Better Few-shot Learners (Gao et al. 2020) <https://arxiv.org/pdf/2012.15723.pdf>`_

    """
    dataset_project = {"train": "train",
                        "dev": "train",
                        "test": "dev"
                        }
    def __init__(self):
        super().__init__()
        self.labels = ['0', '1']
    
    def get_examples(self, data_dir, split):
        path = os.path.join(data_dir, f"{self.dataset_project[split]}.tsv")
        examples = []
        with open(path, encoding='utf-8')as f:
            lines = f.readlines()
            for idx, line in enumerate(lines[1:]):
                linelist = line.strip().split('\t')
                text_a = linelist[0]
                label = linelist[1]
                guid = "%s-%s" % (split, idx)
                example = InputExample(guid=guid, text_a=text_a, label=self.get_label_id(label))
                examples.append(example)
        return examples


class SnliProcessor(DataProcessor):