gpt2-medium-persian / src /create_dataset.py
m3hrdadfi's picture
Add dataset creation script
c92ce97
raw
history blame
4.34 kB
import ast
import logging
import os
import sys
from dataclasses import dataclass, field
import pandas as pd
from sklearn.model_selection import train_test_split
from tqdm import tqdm
from typing import Dict, List, Optional, Tuple
from datasets import load_dataset
from transformers import (
HfArgumentParser,
)
from data_utils import (
filter_by_lang_regex,
filter_by_num_tokens,
filter_by_num_sents,
filter_by_adv,
normalizer
)
logger = logging.getLogger(__name__)
@dataclass
class DataArguments:
"""
Arguments to which dataset we are going to set up.
"""
output_dir: str = field(
default=".",
metadata={"help": "The output directory where the config will be written."},
)
dataset_name: str = field(
default=None,
metadata={"help": "The name of the dataset to use (via the datasets library)."}
)
dataset_config_name: Optional[str] = field(
default=None, metadata={"help": "The configuration name of the dataset to use (via the datasets library)."}
)
train_file: Optional[str] = field(default=None, metadata={"help": "The input training data file (a text file)."})
cache_dir: Optional[str] = field(
default=None,
metadata={"help": "Where do you want to store the pretrained models downloaded from huggingface.co"},
)
def main():
parser = HfArgumentParser([DataArguments])
if len(sys.argv) == 2 and sys.argv[1].endswith(".json"):
# If we pass only one argument to the script and it's the path to a json file,
# let's parse it to get our arguments.
data_args = parser.parse_json_file(json_file=os.path.abspath(sys.argv[1]))[0]
else:
data_args = parser.parse_args_into_dataclasses()[0]
# Setup logging
logging.basicConfig(
format="%(asctime)s - %(levelname)s - %(name)s - %(message)s",
datefmt="%m/%d/%Y %H:%M:%S",
handlers=[logging.StreamHandler(sys.stdout)],
)
logger.setLevel(logging.INFO)
logger.info(f"Preparing the dataset")
if data_args.dataset_name is not None:
dataset = load_dataset(
data_args.dataset_name,
data_args.dataset_config_name,
cache_dir=data_args.cache_dir,
split="train"
)
else:
data_files = {"train": data_args.train_file}
extension = data_args.train_file.split(".")[-1]
if extension == "txt":
extension = "text"
dataset = load_dataset(
extension,
data_files=data_files,
delimiter="\t",
cache_dir=data_args.cache_dir,
)
logger.info(f"dataset: {dataset}")
def data_preparation(item_dict):
if "text" not in item_dict:
return None
text = item_dict["text"]
status = filter_by_lang_regex(text, ratio=0.75)
if not status:
return None
status = filter_by_num_tokens(text, gt=64)
if not status:
return None
status = filter_by_num_sents(text, gt=2)
if not status:
return None
status = filter_by_adv(text, ratio=50)
if not status:
return None
text = normalizer(text)
return {"text": text}
data_dict = []
for item in tqdm(dataset, position=0, total=len(dataset)):
item = data_preparation(item)
if item:
data_dict.append(item)
data_df = pd.DataFrame(data_dict)
logger.info(f"Preparation - [before] consists of {len(dataset)} records!")
logger.info(f"Preparation - [after] consists of {len(data_df)} records!")
train, test = train_test_split(data_df, test_size=0.01, random_state=101)
train = train.reset_index(drop=True)
test = test.reset_index(drop=True)
logger.info(f"Preparation of [train] set consists of {len(train)} records!")
logger.info(f"Preparation of [test] set consists of {len(test)} records!")
os.makedirs(data_args.output_dir, exist_ok=True)
train.to_csv(os.path.join(data_args.output_dir, "train.csv"), sep="\t", encoding="utf-8", index=False)
test.to_csv(os.path.join(data_args.output_dir, "test.csv"), sep="\t", encoding="utf-8", index=False)
logger.info(f"Data saved here {data_args.output_dir}")
if __name__ == '__main__':
main()