metadata
configs:
- config_name: single_256
data_files:
- path: single_256/data-00000-of-00001.arrow
split: train
- config_name: single_512
data_files:
- path: single_512/data-00000-of-00001.arrow
split: train
- config_name: single_1024
data_files:
- path: single_1024/data-00000-of-00001.arrow
split: train
- config_name: single_2048
data_files:
- path: single_2048/data-00000-of-00001.arrow
split: train
- config_name: single_4096
data_files:
- path: single_4096/data-00000-of-00001.arrow
split: train
- config_name: multiple_256
data_files:
- path: multiple_256/data-00000-of-00001.arrow
split: train
- config_name: multiple_512
data_files:
- path: multiple_512/data-00000-of-00001.arrow
split: train
- config_name: multiple_1024
data_files:
- path: multiple_1024/data-00000-of-00001.arrow
split: train
- config_name: multiple_2048
data_files:
- path: multiple_2048/data-00000-of-00001.arrow
split: train
- config_name: multiple_4096
data_files:
- path: multiple_4096/data-00000-of-00001.arrow
split: train
- config_name: delete_256
data_files:
- path: delete_256/data-00000-of-00001.arrow
split: train
- config_name: delete_512
data_files:
- path: delete_512/data-00000-of-00001.arrow
split: train
- config_name: delete_1024
data_files:
- path: delete_1024/data-00000-of-00001.arrow
split: train
- config_name: delete_2048
data_files:
- path: delete_2048/data-00000-of-00001.arrow
split: train
- config_name: delete_4096
data_files:
- path: delete_4096/data-00000-of-00001.arrow
split: train
Dataset Organization
The DEG dataset is organized by subsets (single, multiple, delete) and context length. Each subset has multiple subsets corresponding to different context lengths.
Instructions to Download
You can load a dataset using HF's API, with an example below.
from datasets import load_dataset
deg_data = load_dataset("wanglab/deg", 'single_256')
# Access the train data for a specific context length
train_data = deg_data['train']