metadata
dataset_info:
features:
- name: audio
dtype: audio
- name: id
dtype: int64
- name: brand_id
dtype: int64
- name: brand_name
dtype: string
- name: text
dtype: string
- name: title
dtype: string
- name: created_at
dtype: timestamp[us, tz=UTC]
- name: confirmed_at
dtype: timestamp[us, tz=UTC]
- name: confirmed_by_id
dtype: int64
- name: clip_url
dtype: string
- name: duration
dtype: float64
- name: thumb_url
dtype: string
- name: clip_duration
dtype: float64
- name: filename
dtype: string
- name: embeddings
sequence:
sequence: float32
splits:
- name: train
num_bytes: 261559300
num_examples: 417
download_size: 242934514
dataset_size: 261559300
configs:
- config_name: default
data_files:
- split: train
path: data/train-*