README.md

Overview

Original dataset page here and dataset available here.

Dataset curation

Added new column label with encoded labels with the following mapping

{"entailment": 0, "neutral": 1, "contradiction": 2}

and the columns with parse information are dropped as they are not well formatted.

Also, the name of the file from which each instance comes is added in the column dtype.

Code to create the dataset

import pandas as pd
from datasets import Dataset, ClassLabel, Value, Features, DatasetDict
import json
from pathlib import Path


# load data
ds = {}
path = Path("<path to folder>")
for i in path.rglob("*.jsonl"): 
    print(i)
    name = str(i).split("/")[0].lower()
    dtype = str(i).split("/")[1].lower()
    
    # read data
    with i.open("r") as fl:
        df = pd.DataFrame([json.loads(line) for line in fl])
    
    # select columns
    df = df.loc[:, ["sentence1", "sentence2", "gold_label"]]
    
    # add file name as column
    df["dtype"] = dtype
    
    # encode labels
    df["label"] = df["gold_label"].map({"entailment": 0, "neutral": 1, "contradiction": 2})
    ds[name] = df
        
# cast to dataset
features = Features(
    {
        "sentence1": Value(dtype="string"),
        "sentence2": Value(dtype="string"),
        "label": ClassLabel(num_classes=3, names=["entailment", "neutral", "contradiction"]),
        "dtype": Value(dtype="string"),
        "gold_label": Value(dtype="string"),
    }
)
ds = DatasetDict({k: Dataset.from_pandas(v, features=features) for k, v in ds.items()})
ds.push_to_hub("pietrolesci/stress_tests_nli", token="<token>")


# check overlap between splits
from itertools import combinations
for i, j in combinations(ds.keys(), 2):
    print(
        f"{i} - {j}: ",
        pd.merge(
            ds[i].to_pandas(), 
            ds[j].to_pandas(), 
            on=["sentence1", "sentence2", "label"], 
            how="inner",
        ).shape[0],
    )
#> numerical_reasoning - negation:  0
#> numerical_reasoning - length_mismatch:  0
#> numerical_reasoning - spelling_error:  0
#> numerical_reasoning - word_overlap:  0
#> numerical_reasoning - antonym:  0
#> negation - length_mismatch:  0
#> negation - spelling_error:  0
#> negation - word_overlap:  0
#> negation - antonym:  0
#> length_mismatch - spelling_error:  0
#> length_mismatch - word_overlap:  0
#> length_mismatch - antonym:  0
#> spelling_error - word_overlap:  0
#> spelling_error - antonym:  0
#> word_overlap - antonym:  0