metadata
dataset_info:
features:
- name: run_name
dtype: string
- name: checkpoint
dtype: int64
- name: task
dtype: string
- name: pooling
dtype: string
- name: score
dtype: float64
- name: autoencoding_question
dtype: bool
- name: autoencoding_answer
dtype: bool
- name: generative_question_to_answer
dtype: bool
- name: generative_answer_to_question
dtype: bool
- name: autoencoding_question_weight
dtype: float64
- name: autoencoding_answer_weight
dtype: float64
- name: generative_question_to_answer_weight
dtype: float64
- name: generative_answer_to_question_weight
dtype: float64
- name: __index_level_0__
dtype: int64
splits:
- name: train
num_bytes: 116952
num_examples: 1020
download_size: 30905
dataset_size: 116952
configs:
- config_name: default
data_files:
- split: train
path: data/train-*