metadata
dataset_info:
- config_name: boolq
features:
- name: dataset
dtype: string
- name: kind
dtype: string
- name: context
dtype: string
- name: raw_question
dtype: string
- name: full_question
dtype: string
- name: gold
dtype: string
- name: gold_refs
list: 'null'
- name: answer_type
dtype: string
- name: is_mc
dtype: bool
- name: num_options
dtype: int64
- name: generated_reasoning
dtype: string
- name: generated_answer
dtype: string
- name: concise_answer
dtype: string
- name: answer_matches_source
dtype: bool
- name: reward
dtype: float64
- name: reasoning_effort
dtype: string
- name: finish_reason
dtype: string
splits:
- name: train
num_bytes: 25412377
num_examples: 7864
download_size: 11727466
dataset_size: 25412377
- config_name: dream
features:
- name: dataset
dtype: string
- name: kind
dtype: string
- name: context
dtype: string
- name: raw_question
dtype: string
- name: full_question
dtype: string
- name: gold
dtype: string
- name: gold_refs
list: 'null'
- name: answer_type
dtype: string
- name: is_mc
dtype: bool
- name: num_options
dtype: int64
- name: generated_reasoning
dtype: string
- name: generated_answer
dtype: string
- name: concise_answer
dtype: string
- name: answer_matches_source
dtype: bool
- name: reward
dtype: float64
- name: reasoning_effort
dtype: string
- name: finish_reason
dtype: string
splits:
- name: train
num_bytes: 17520482
num_examples: 5650
download_size: 6455302
dataset_size: 17520482
- config_name: drop
features:
- name: dataset
dtype: string
- name: kind
dtype: string
- name: context
dtype: string
- name: raw_question
dtype: string
- name: full_question
dtype: string
- name: gold
dtype: string
- name: gold_refs
list: 'null'
- name: answer_type
dtype: string
- name: is_mc
dtype: bool
- name: num_options
dtype: int64
- name: generated_reasoning
dtype: string
- name: generated_answer
dtype: string
- name: concise_answer
dtype: string
- name: answer_matches_source
dtype: bool
- name: reward
dtype: float64
- name: reasoning_effort
dtype: string
- name: finish_reason
dtype: string
splits:
- name: train
num_bytes: 146367015
num_examples: 30632
download_size: 38685245
dataset_size: 146367015
- config_name: pubmedqa
features:
- name: dataset
dtype: string
- name: kind
dtype: string
- name: context
dtype: string
- name: raw_question
dtype: string
- name: full_question
dtype: string
- name: gold
dtype: string
- name: gold_refs
list: 'null'
- name: answer_type
dtype: string
- name: is_mc
dtype: bool
- name: num_options
dtype: int64
- name: generated_reasoning
dtype: string
- name: generated_answer
dtype: string
- name: concise_answer
dtype: string
- name: answer_matches_source
dtype: bool
- name: reward
dtype: float64
- name: reasoning_effort
dtype: string
- name: finish_reason
dtype: string
splits:
- name: train
num_bytes: 212826296
num_examples: 48050
download_size: 108331725
dataset_size: 212826296
- config_name: quoref
features:
- name: dataset
dtype: string
- name: kind
dtype: string
- name: context
dtype: string
- name: raw_question
dtype: string
- name: full_question
dtype: string
- name: gold
dtype: string
- name: gold_refs
list: 'null'
- name: answer_type
dtype: string
- name: is_mc
dtype: bool
- name: num_options
dtype: int64
- name: generated_reasoning
dtype: string
- name: generated_answer
dtype: string
- name: concise_answer
dtype: string
- name: answer_matches_source
dtype: bool
- name: reward
dtype: float64
- name: reasoning_effort
dtype: string
- name: finish_reason
dtype: string
splits:
- name: train
num_bytes: 76270110
num_examples: 11849
download_size: 28999875
dataset_size: 76270110
- config_name: race
features:
- name: dataset
dtype: string
- name: kind
dtype: string
- name: context
dtype: string
- name: raw_question
dtype: string
- name: full_question
dtype: string
- name: gold
dtype: string
- name: gold_refs
list: 'null'
- name: answer_type
dtype: string
- name: is_mc
dtype: bool
- name: num_options
dtype: int64
- name: generated_reasoning
dtype: string
- name: generated_answer
dtype: string
- name: concise_answer
dtype: string
- name: answer_matches_source
dtype: bool
- name: reward
dtype: float64
- name: reasoning_effort
dtype: string
- name: finish_reason
dtype: string
splits:
- name: train
num_bytes: 242538361
num_examples: 42963
download_size: 103469254
dataset_size: 242538361
- config_name: sciriff-yesno
features:
- name: dataset
dtype: string
- name: kind
dtype: string
- name: context
dtype: string
- name: raw_question
dtype: string
- name: full_question
dtype: string
- name: gold
dtype: string
- name: gold_refs
list: 'null'
- name: answer_type
dtype: string
- name: is_mc
dtype: bool
- name: num_options
dtype: int64
- name: generated_reasoning
dtype: string
- name: generated_answer
dtype: string
- name: concise_answer
dtype: string
- name: answer_matches_source
dtype: bool
- name: reward
dtype: float64
- name: reasoning_effort
dtype: string
- name: finish_reason
dtype: string
splits:
- name: train
num_bytes: 7082137
num_examples: 1372
download_size: 3382083
dataset_size: 7082137
- config_name: squad_v2
features:
- name: dataset
dtype: string
- name: kind
dtype: string
- name: context
dtype: string
- name: raw_question
dtype: string
- name: full_question
dtype: string
- name: gold
dtype: string
- name: gold_refs
list: 'null'
- name: answer_type
dtype: string
- name: is_mc
dtype: bool
- name: num_options
dtype: int64
- name: generated_reasoning
dtype: string
- name: generated_answer
dtype: string
- name: concise_answer
dtype: string
- name: answer_matches_source
dtype: bool
- name: reward
dtype: float64
- name: reasoning_effort
dtype: string
- name: finish_reason
dtype: string
splits:
- name: train
num_bytes: 88571701
num_examples: 27602
download_size: 32650194
dataset_size: 88571701
- config_name: tabfact
features:
- name: dataset
dtype: string
- name: kind
dtype: string
- name: context
dtype: string
- name: raw_question
dtype: string
- name: full_question
dtype: string
- name: gold
dtype: string
- name: gold_refs
list: 'null'
- name: answer_type
dtype: string
- name: is_mc
dtype: bool
- name: num_options
dtype: int64
- name: generated_reasoning
dtype: string
- name: generated_answer
dtype: string
- name: concise_answer
dtype: string
- name: answer_matches_source
dtype: bool
- name: reward
dtype: float64
- name: reasoning_effort
dtype: string
- name: finish_reason
dtype: string
splits:
- name: train
num_bytes: 186829043
num_examples: 42908
download_size: 60948720
dataset_size: 186829043
- config_name: vitaminc
features:
- name: dataset
dtype: string
- name: kind
dtype: string
- name: context
dtype: string
- name: raw_question
dtype: string
- name: full_question
dtype: string
- name: gold
dtype: string
- name: gold_refs
list: 'null'
- name: answer_type
dtype: string
- name: is_mc
dtype: bool
- name: num_options
dtype: int64
- name: generated_reasoning
dtype: string
- name: generated_answer
dtype: string
- name: concise_answer
dtype: string
- name: answer_matches_source
dtype: bool
- name: reward
dtype: float64
- name: reasoning_effort
dtype: string
- name: finish_reason
dtype: string
splits:
- name: train
num_bytes: 94259452
num_examples: 37953
download_size: 27774320
dataset_size: 94259452
- config_name: wtq
features:
- name: dataset
dtype: string
- name: kind
dtype: string
- name: context
dtype: string
- name: raw_question
dtype: string
- name: full_question
dtype: string
- name: gold
dtype: string
- name: gold_refs
list: string
- name: answer_type
dtype: string
- name: is_mc
dtype: bool
- name: num_options
dtype: int64
- name: generated_reasoning
dtype: string
- name: generated_answer
dtype: string
- name: concise_answer
dtype: string
- name: answer_matches_source
dtype: bool
- name: reward
dtype: float64
- name: reasoning_effort
dtype: string
- name: finish_reason
dtype: string
splits:
- name: train
num_bytes: 61984018
num_examples: 11537
download_size: 27603256
dataset_size: 61984018
configs:
- config_name: boolq
data_files:
- split: train
path: boolq/train-*
- config_name: dream
data_files:
- split: train
path: dream/train-*
- config_name: drop
data_files:
- split: train
path: drop/train-*
- config_name: pubmedqa
data_files:
- split: train
path: pubmedqa/train-*
- config_name: quoref
data_files:
- split: train
path: quoref/train-*
- config_name: race
data_files:
- split: train
path: race/train-*
- config_name: sciriff-yesno
data_files:
- split: train
path: sciriff-yesno/train-*
- config_name: squad_v2
data_files:
- split: train
path: squad_v2/train-*
- config_name: tabfact
data_files:
- split: train
path: tabfact/train-*
- config_name: vitaminc
data_files:
- split: train
path: vitaminc/train-*
- config_name: wtq
data_files:
- split: train
path: wtq/train-*