metadata
dataset_info:
features:
- name: qwen2-72b
dtype: string
- name: llama3-8b-it-new
dtype: string
- name: arxiv-translator
dtype: string
- name: deepl
dtype: string
- name: llama3-instrucTrans-enko-8b
dtype: string
- name: llama3.1-8b-it-new
dtype: string
splits:
- name: train
num_bytes: 3829555
num_examples: 950
download_size: 2012247
dataset_size: 3829555
configs:
- config_name: default
data_files:
- split: train
path: data/train-*