metadata
dataset_info:
features:
- name: text
dtype: string
- name: meta
struct:
- name: pile_set_name
dtype: string
splits:
- name: train
num_bytes: 436655214
num_examples: 80000
- name: validation
num_bytes: 57811151
num_examples: 10000
- name: test
num_bytes: 59411825
num_examples: 10000
download_size: 289955074
dataset_size: 553878190
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*
Copy of jbrinkma/pile-100k. This version split into: 80k train, 10k test, 10k validation
Features of this version
- This version has a both a train + test + validation set
- Is easily downloadable in ~300MB
- Keeps metadata