metadata
dataset_info:
features:
- name: question
dtype: string
- name: answer
dtype: string
- name: task_config
dtype: string
- name: task_source
dtype: string
- name: prompt
list:
- name: content
dtype: string
- name: role
dtype: string
- name: model_responses
sequence: 'null'
- name: model_responses__eval_is_correct
sequence: 'null'
- name: all_other_columns
dtype: string
- name: original_split
dtype: string
- name: metadata
dtype: string
- name: model_responses__best_of_n_atags
sequence: string
- name: model_responses__best_of_n_atags__finish_reason_length_flags
sequence: bool
- name: model_responses__best_of_n_atags__length_partial_responses
sequence: string
- name: prompt__best_of_n_atags__metadata
dtype: string
- name: model_responses__best_of_n_atags__metadata
dtype: string
- name: model_responses__best_of_n_atags__eval_is_correct
sequence: bool
- name: model_responses__best_of_n_atags__eval_extracted_answers
sequence: string
- name: model_responses__best_of_n_atags__eval_extraction_metadata
dtype: string
- name: model_responses__best_of_n_atags__eval_evaluation_metadata
dtype: string
- name: model_responses__best_of_n_atags__internal_answers__eval_is_correct
sequence:
sequence: bool
- name: >-
model_responses__best_of_n_atags__internal_answers__eval_extracted_answers
sequence:
sequence: string
- name: >-
model_responses__best_of_n_atags__internal_answers__eval_extraction_metadata
dtype: string
- name: >-
model_responses__best_of_n_atags__internal_answers__eval_evaluation_metadata
dtype: string
- name: model_responses__best_of_n_atags__metrics
struct:
- name: flips_by
sequence: int64
- name: flips_total
dtype: int64
- name: num_correct
dtype: int64
- name: pass_at_n
dtype: int64
- name: percent_correct
dtype: float64
- name: skill_count
struct:
- name: answer_revision
sequence: int64
- name: best_of_n
sequence: int64
- name: reflect_close
sequence: int64
- name: reflect_open
sequence: int64
- name: reflection_sbon
sequence: int64
- name: sample_close
sequence: int64
- name: sample_open
sequence: int64
- name: vote_close
sequence: int64
- name: vote_open
sequence: int64
- name: voting
sequence: int64
- name: total_responses
dtype: int64
- name: eval_date
dtype: string
splits:
- name: train
num_bytes: 2116596621
num_examples: 4000
download_size: 874383255
dataset_size: 2116596621
configs:
- config_name: default
data_files:
- split: train
path: data/train-*