nli-zh-all.py

# coding=utf-8
# Copyright 2020 The TensorFlow Datasets Authors and the HuggingFace Datasets Authors.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

# Lint as: python3
"""The Chinese Natural Language Inference (NLI-zh-all) Corpus.
upload： https://github.com/shibing624
"""


import csv
import os
import json
import datasets


_CITATION = """https://github.com/shibing624/text2vec"""

_DESCRIPTION = """\
The SNLI corpus (version 1.0) is a merged chinese sentence similarity dataset, supporting the task of natural language
inference (NLI), also known as recognizing textual entailment (RTE).
"""

_DATA_URL = "https://huggingface.co/datasets/shibing624/nli-zh-all/resolve/main/sampled_data"


class Nli(datasets.GeneratorBasedBuilder):
    """The Chinese Natural Language Inference (NLI-zh-all) Corpus."""

    BUILDER_CONFIGS = [
        datasets.BuilderConfig(
            name="plain_text",
            version=datasets.Version("1.0.0", ""),
            description="Plain text import of NLI-zh-all",
        )
    ]

    def _info(self):
        return datasets.DatasetInfo(
            description=_DESCRIPTION,
            features=datasets.Features(
                {
                    "text1": datasets.Value("string"),
                    "text2": datasets.Value("string"),
                    "label": datasets.Value("int64"),
                }
            ),

            supervised_keys=None,
            homepage="https://github.com/shibing624/text2vec",
            citation=_CITATION,
        )

    def _split_generators(self, dl_manager):
        
        files = ['simclue-train-2k.jsonl',
                'nli_zh-train-25k.jsonl',
                'alpaca_gpt4-train-2k.jsonl',
                'cmrc2018-train-2k.jsonl',
                'snli_zh-train-5k.jsonl',
                'chatmed_consult-train-500.jsonl',
                'zhihu_kol-train-2k.jsonl',
                'cblue_chip_sts-train-2k.jsonl',
                'csl-train-500.jsonl',
                'webqa-train-500.jsonl',
                'xlsum-train-1k.jsonl',]
        data_files = [f"{_DATA_URL}/{i}" for i in files]
        return [
            
            datasets.SplitGenerator(
                name=datasets.Split.TRAIN, gen_kwargs={"filepath": dl_manager.download_and_extract(data_files)}
            ),
        ]

    def _generate_examples(self, filepath):
        """This function returns the examples in the raw (text) form."""
        id = 0
        if isinstance(filepath, str):
            filepath = [filepath]
        for file in filepath:
            with open(file, encoding="utf-8") as f:
                for key, row in enumerate(f):
                    data = json.loads(row)
                    yield id, {
                            "text1": data["text1"],
                            "text2": data["text2"],
                            "label": data["label"]
                    }
                    id += 1