File size: 3,244 Bytes
66407c5 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 |
# Copyright 2024 Bytedance Ltd. and/or its affiliates
#
# 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.
"""
Preprocess Hellaswag dataset.
"""
import argparse
import os
import re
import datasets
from verl.utils.hdfs_io import copy, makedirs
def preprocess(text):
text = text.strip()
# NOTE: Brackets are artifacts of the WikiHow dataset portion of HellaSwag.
text = text.replace(" [title]", ". ")
text = re.sub("\\[.*?\\]", "", text)
text = text.replace(" ", " ")
return text
if __name__ == "__main__":
parser = argparse.ArgumentParser()
parser.add_argument("--local_dir", default="/opt/tiger/hellaswag")
parser.add_argument("--hdfs_dir", default=None)
args = parser.parse_args()
data_source = "Rowan/hellaswag"
dataset = datasets.load_dataset(data_source, trust_remote_code=True)
train_dataset = dataset["train"]
val_dataset = dataset["validation"]
test_dataset = dataset["test"]
instruction = "Please complete the following sentence.\n"
def make_map_fn(split):
def process_fn(doc, idx):
ctx = doc["ctx_a"] + " " + doc["ctx_b"].capitalize()
query = preprocess(doc["activity_label"] + ": " + ctx)
choices = [preprocess(ending) for ending in doc["endings"]]
gold = int(doc["label"])
data = {
"data_source": data_source,
"prompt": [{"role": "user", "content": query}],
"ability": "nlp",
"reward_model": {
"style": "model",
"eval": "multiple_choice", # using loglikelihood
"ground_truth": gold,
"choices": choices,
},
"extra_info": {"split": split, "index": idx},
}
return data
return process_fn
# filter data that doesn't have a label
train_dataset = train_dataset.filter(lambda x: len(x["label"]) > 0)
val_dataset = val_dataset.filter(lambda x: len(x["label"]) > 0)
test_dataset = test_dataset.filter(lambda x: len(x["label"]) > 0)
train_dataset = train_dataset.map(function=make_map_fn("train"), with_indices=True)
val_dataset = val_dataset.map(function=make_map_fn("validation"), with_indices=True)
test_dataset = test_dataset.map(function=make_map_fn("test"), with_indices=True)
local_dir = args.local_dir
hdfs_dir = args.hdfs_dir
train_dataset.to_parquet(os.path.join(local_dir, "train.parquet"))
val_dataset.to_parquet(os.path.join(local_dir, "validation.parquet"))
test_dataset.to_parquet(os.path.join(local_dir, "test.parquet"))
if hdfs_dir is not None:
makedirs(hdfs_dir)
copy(src=local_dir, dst=hdfs_dir)
|