metadata
dataset_info:
features:
- name: uniprot_accession
dtype: string
- name: chain_id
dtype: string
- name: seq
dtype: string
- name: sst3
dtype: string
- name: sst8
dtype: string
- name: len
dtype: int64
- name: input_ids
sequence: int32
- name: attention_mask
sequence: int8
- name: special_tokens_mask
sequence: int8
- name: label
sequence: int64
splits:
- name: train
num_bytes: 897088
num_examples: 99
download_size: 78108
dataset_size: 897088
configs:
- config_name: default
data_files:
- split: train
path: data/train-*