README.md

metadata

dataset_info:
  features:
    - name: title
      dtype: string
    - name: summary
      dtype: string
    - name: link
      dtype: string
    - name: segments
      struct:
        - name: language
          dtype: string
        - name: segments
          list:
            - name: avg_logprob
              dtype: float64
            - name: compression_ratio
              dtype: float64
            - name: end
              dtype: float64
            - name: id
              dtype: int64
            - name: no_speech_prob
              dtype: float64
            - name: seek
              dtype: int64
            - name: start
              dtype: float64
            - name: temperature
              dtype: float64
            - name: text
              dtype: string
            - name: tokens
              sequence: int64
        - name: text
          dtype: string
    - name: transcript
      dtype: string
  splits:
    - name: train
      num_bytes: 11839153
      num_examples: 39
  download_size: 4942577
  dataset_size: 11839153
task_categories:
  - text-classification
  - text-generation
  - summarization
language:
  - en
size_categories:
  - n<1K
pretty_name: TalkRL Podcast

Dataset Card for "talkrl-podcast"

This dataset is sourced from the TalkRL Podcast website and contains English transcripts of wonderful TalkRL podcast episodes. The transcripts were generated using OpenAI's base Whisper model