Invalsi.py

# Copyright 2020 The HuggingFace Datasets Authors and the current dataset script contributor.
#
# 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.
# TODO: Address all TODOs and remove all explanatory comments


import csv
import json
import os

import datasets


_CITATION = """\
@misc{esuli2024invalsi,
    title={The Invalsi Benchmark: measuring Language Models Mathematical and Language understanding in Italian}, 
    author={Andrea Esuli and Giovanni Puccetti},
    year={2024},
    eprint={2403.18697},
    archivePrefix={arXiv},
    primaryClass={cs.CL}
}
"""

_DESCRIPTION = """\
This new dataset is designed to measure Language Models mathematical and language understanding in Italian.
"""

_HOMEPAGE = ""

_LICENSE = "CC BY 4.0"


_URLS = {
    # "mate": "https://huggingface.co/datasets/ai4text/Invalsi/blob/main/invalsi_ita_data.zip",
    # "ita": "https://huggingface.co/datasets/ai4text/Invalsi/blob/main/invalsi_mate_data.zip",
    "mate": "./invalsi_mate_data/invalsi_mate_clean.csv",
    "ita": "./invalsi_ita_data/invalsi_ita_clean.csv",
}


class invalsi(datasets.GeneratorBasedBuilder):

    VERSION = datasets.Version("0.1.0")

    BUILDER_CONFIGS = [
        datasets.BuilderConfig(name="mate", version=VERSION, description="Mathematical Understanding"),
        datasets.BuilderConfig(name="ita", version=VERSION, description="Italian Understanding"),
    ]

    DEFAULT_CONFIG_NAME = "mate"

    def _info(self):
        if self.config.name == "mate":
            features = datasets.Features(
                {
                    "domanda": datasets.Value("string"),
                    "risposta": datasets.Value("string"),
                    "immagine": datasets.Value("string"),
                    "test_id": datasets.Value("string"),
                }
            )
        elif self.config.name == "ita":
            features = datasets.Features(
                {
                    "testo": datasets.Value("string"),
                    "domanda": datasets.Value("string"),
                    "risposta": datasets.Value("string"),
                    "immagine": datasets.Value("string"),
                    "test_id": datasets.Value("string"),
                }
            )

        return datasets.DatasetInfo(
            description=_DESCRIPTION,
            features=features,
            homepage=_HOMEPAGE,
            license=_LICENSE,
            citation=_CITATION,
        )

    def _split_generators(self, dl_manager):
        # TODO: This method is tasked with downloading/extracting the data and defining the splits depending on the configuration
        # If several configurations are possible (listed in BUILDER_CONFIGS), the configuration selected by the user is in self.config.name

        # dl_manager is a datasets.download.DownloadManager that can be used to download and extract URLS
        # It can accept any type or nested list/dict and will give back the same structure with the url replaced with path to local files.
        # By default the archives will be extracted and a path to a cached folder where they are extracted is returned instead of the archive
        urls = _URLS[self.config.name]
        # data_dir = dl_manager.download_and_extract(urls)
        data_dir = "./"
        if self.config.name == "mate":
            data_file = "invalsi_mate_data/invalsi_mate_clean.csv"
        elif self.config.name == "ita":
            data_file = "invalsi_ita_data/invalsi_ita_clean.csv"
        return [
            datasets.SplitGenerator(
                name=datasets.Split.VALIDATION,
                # These kwargs will be passed to _generate_examples
                gen_kwargs={
                    "filepath": os.path.join(data_dir, data_file),
                    "split": "val",
                },
            ),
        ]

    def _generate_examples(self, filepath, split):
        ds = datasets.load_dataset("csv", data_files=filepath)["train"]
        for key, row in enumerate(ds):
            # data = json.loads(row)
            if self.config.name == "mate":
                # Yields examples as (key, example) tuples
                out = {
                    # "domanda": datasets.Value("string"),
                    # "risposta": datasets.Value("string"),
                    # "immagine": datasets.Value("string"),
                    # "test_id": datasets.Value("string"),
                    "domanda": row["domanda"],
                    "risposta": row["risposta"],

                    "test_id": row["test_id"],
                }
                if "image_file_names" in row:
                    out["immagine"] = row["image_file_names"]

                yield key, out
            elif self.config.name == "ita":
                yield key, {
                    # "testo": datasets.Value("string"),
                    # "domanda": datasets.Value("string"),
                    # "risposta": datasets.Value("string"),
                    # "immagine": datasets.Value("string"),
                    # "test_id": datasets.Value("string"),
                    "testo": row["testo"],
                    "domanda": row["domanda"],
                    "risposta": row["risposta"],
                    "immagine": row["image_file_names"],
                    "test_id": row["test_id"],
                }