README.md

---
dataset_info:
  features:
  - name: pitched_product
    dtype: string
  - name: pitched_brand
    dtype: string
  - name: pitched_industry
    dtype: string
  - name: pitched_description
    dtype: string
  - name: video_text
    dtype: string
  - name: bucket
    dtype: string
  - name: brand
    dtype: string
  - name: ad_style
    dtype: string
  - name: has_promo_tie_in
    dtype: bool
  - name: anomaly_hint
    dtype: bool
  - name: anomaly_reason
    dtype: string
  splits:
  - name: train
    num_bytes: 17364021
    num_examples: 11058
  - name: test
    num_bytes: 1929859
    num_examples: 1229
  download_size: 8892762
  dataset_size: 19293880
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---