metadata
dataset_info:
features:
- name: context
dtype: string
- name: question
dtype: string
- name: answers
struct:
- name: answer_start
sequence: int64
- name: text
sequence: string
- name: id
dtype: string
- name: labels
list:
- name: end
sequence: int64
- name: start
sequence: int64
splits:
- name: train
num_bytes: 57635506.94441748
num_examples: 18142
- name: validation
num_bytes: 3374870.9449192784
num_examples: 1070
download_size: 4666280
dataset_size: 61010377.88933676
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
Dataset Card for "squad"
This truncated dataset is derived from the Stanford Question Answering Dataset (SQuAD) for reading comprehension. Its primary aim is to extract instances from the original SQuAD dataset that align with the context length of BERT, RoBERTa, OPT, and T5 models.
Preprocessing and Filtering
Preprocessing involves tokenization using the BertTokenizer (WordPiece), RoBertaTokenizer (Byte-level BPE), OPTTokenizer (Byte-Pair Encoding), and T5Tokenizer (Sentence Piece). Each sample is then checked to ensure that the length of the tokenized input is within the specified model_max_length for each tokenizer.