metadata
dataset_info:
features:
- name: instruction_seed
dtype: string
- name: response_seed
dtype: string
- name: final_reasoning_trace_gemini
dtype: string
- name: __original_row_idx
dtype: int64
splits:
- name: train
num_bytes: 3223512
num_examples: 300
download_size: 1521386
dataset_size: 3223512
configs:
- config_name: default
data_files:
- split: train
path: data/train-*