metadata
dataset_info:
features:
- name: chanel_name
dtype: string
- name: video_id
dtype: string
- name: source
dtype: string
- name: publish_date
dtype: string
- name: duration
dtype: float64
- name: last_update_date
dtype: string
- name: title
dtype: string
- name: text
dtype: string
- name: year
dtype: int64
- name: number_of_tokenks
dtype: int64
- name: number_of_new_tokens
dtype: float64
- name: prompt
dtype: string
- name: summary
dtype: string
- name: key_terms
sequence: string
- name: __index_level_0__
dtype: int64
- name: summary_token_length
dtype: int64
- name: summary_llama3_1_8b
dtype: string
splits:
- name: train
num_bytes: 23776369
num_examples: 622
download_size: 11870840
dataset_size: 23776369
configs:
- config_name: default
data_files:
- split: train
path: data/train-*