README.md

metadata

license: mit
dataset_info:
  features:
    - name: nmId
      dtype: int64
    - name: text
      dtype: string
    - name: responder
      dtype: string
    - name: type
      dtype: string
    - name: product_name
      dtype: string
    - name: product_category_2
      dtype: string
    - name: product_category_1
      dtype: string
    - name: product_color
      dtype: string
    - name: product_description
      dtype: string
    - name: product_brand
      dtype: string
    - name: vector
      sequence: float32
    - name: toxicity
      dtype: string
    - name: emotions
      dtype: string
    - name: target_nmId
      dtype: int64
    - name: target_text
      dtype: string
    - name: target_responder
      dtype: string
    - name: target_type
      dtype: string
    - name: target_product_name
      dtype: string
    - name: target_product_category_2
      dtype: string
    - name: target_product_category_1
      dtype: string
    - name: target_product_color
      dtype: string
    - name: target_product_description
      dtype: string
    - name: target_product_brand
      dtype: string
    - name: target_vector
      sequence: float32
    - name: target_toxicity
      dtype: string
    - name: target_emotions
      dtype: string
  splits:
    - name: train
      num_bytes: 154839107
      num_examples: 22383
  download_size: 138553226
  dataset_size: 154839107
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
task_categories:
  - text2text-generation
  - feature-extraction
  - text-classification
language:
  - ru
size_categories:
  - 10K<n<100K

This is a dataset of questions, answers, feedbacks and products from the Russian marketplace Wildberries. With additional matching pair of textes.

Original data scrapped by nyuuzyou