Source code for

import os
from pathlib import Path
from typing import Union, Tuple

from torchtext._internal.module_utils import is_module_available
from import _create_dataset_directory
from import _wrap_split_argument

if is_module_available("torchdata"):
    from torchdata.datapipes.iter import FileOpener, IterableWrapper, HttpReader

URL = ""

MD5 = "7c2ac02c03563afcf9b574c7e56c153a"

    "train": 25000,
    "test": 25000,

_PATH = "aclImdb_v1.tar.gz"


[docs]@_create_dataset_directory(dataset_name=DATASET_NAME) @_wrap_split_argument(("train", "test")) def IMDB(root: str, split: Union[Tuple[str], str]): """IMDB Dataset For additional details refer to Number of lines per split: - train: 25000 - test: 25000 Args: root: Directory where the datasets are saved. Default: os.path.expanduser('~/.torchtext/cache') split: split or splits to be returned. Can be a string or tuple of strings. Default: (`train`, `test`) :returns: DataPipe that yields tuple of label (1 to 2) and text containing the movie review :rtype: (int, str) """ if not is_module_available("torchdata"): raise ModuleNotFoundError( "Package `torchdata` not found. Please install following instructions at ``" ) url_dp = IterableWrapper([URL]) cache_compressed_dp = url_dp.on_disk_cache( filepath_fn=lambda x: os.path.join(root, _PATH), hash_dict={os.path.join(root, _PATH): MD5}, hash_type="md5", ) cache_compressed_dp = HttpReader(cache_compressed_dp).end_caching( mode="wb", same_filepath_fn=True ) labels = {"neg", "pos"} decompressed_folder = "aclImdb_v1" cache_decompressed_dp = cache_compressed_dp.on_disk_cache( filepath_fn=lambda x: [ os.path.join(root, decompressed_folder, split, label) for label in labels ] ) cache_decompressed_dp = FileOpener(cache_decompressed_dp, mode="b") cache_decompressed_dp = cache_decompressed_dp.read_from_tar() def filter_imdb_data(key, fname): # eg. fname = "aclImdb/train/neg/12416_3.txt" *_, split, label, file = Path(fname).parts return key == split and label in labels cache_decompressed_dp = cache_decompressed_dp.filter( lambda t: filter_imdb_data(split, t[0]) ) # eg. "aclImdb/train/neg/12416_3.txt" -> "neg" cache_decompressed_dp = lambda t: (Path(t[0]).parts[-2], t[1]) ) cache_decompressed_dp = cache_decompressed_dp.readlines(decode=True) cache_decompressed_dp = ( cache_decompressed_dp.lines_to_paragraphs() ) # group by label in cache file cache_decompressed_dp = x: (x[0], x[1].encode())) cache_decompressed_dp = cache_decompressed_dp.end_caching( mode="wb", filepath_fn=lambda x: os.path.join(root, decompressed_folder, split, x), skip_read=True ) # TODO: read in text mode with utf-8 encoding, see: data_dp = FileOpener(cache_decompressed_dp, mode="b") # get label from cache file, eg. "aclImdb_v1/train/neg" -> "neg" return data_dp.readlines(decode=True).map(lambda t: (Path(t[0]).parts[-1], t[1]))


Access comprehensive developer documentation for PyTorch

View Docs


Get in-depth tutorials for beginners and advanced developers

View Tutorials


Find development resources and get your questions answered

View Resources