dataset.py

from datasets import GeneratorBasedBuilder, DatasetInfo, SplitGenerator, DownloadManager
from typing import Any, Dict, List, Tuple
import os
import json

class ProcessedImageDataset(GeneratorBasedBuilder):
    VERSION = "1.0.0" 

    def _info(self) -> DatasetInfo:
        # Specify dataset info here
        return DatasetInfo(
            # You can add description, citation, homepage, etc.
            features=self._features(),
            supervised_keys=("image", "text"),
        )

    def _features(self):
        # Define the features of your dataset: image file, text, etc.
        from datasets import Features, Image, Value
        return Features({"image_file": Image(), "text": Value("string")})

    def _split_generators(self, dl_manager) -> List[SplitGenerator]:
        # This method is tasked with downloading/extracting the data and defining the splits
        print(self.config.data_dir)
        if self.config.data_dir is None:
            raise ValueError(f'Data directory unspecified. Correct usage is: load_dataset(script_path, data_dir=data_dir_path)')
        return [SplitGenerator(name="train", gen_kwargs={"data_dir": self.config.data_dir})]

    def _generate_examples(self, data_dir):
        # def _generate_examples(self, data_dir: str) -> Tuple[int, Dict[str, Any]]:
        # This method will read the data and yield examples
        metadata_file_path = os.path.join(data_dir, "metadata.jsonl")

        # Read metadata and store it in a dictionary
        metadata = {}
        with open(metadata_file_path, "r") as f:
            for line in f:
                item = json.loads(line)
                metadata[item["file_name"]] = item

        # Iterate through each file in the data directory
        for filename in os.listdir(data_dir):
            if filename.endswith(".png") or filename.endswith(".jpg") or filename.endswith(".jpeg"):
                if filename in metadata:
                    metadata_entry = metadata[filename]
                    yield filename, {
                        "image_file": os.path.join(data_dir, filename),
                        "text": metadata_entry["text"],
                    }