README.md

---
dataset_info:
  features:
  - name: episode_id
    dtype: string
  - name: segment_id
    dtype: string
  - name: text
    dtype: string
  - name: prev_text
    dtype: string
  - name: start
    dtype: float64
  - name: end
    dtype: float64
  - name: speaker
    dtype: string
  - name: language
    dtype: string
  - name: detected_language
    dtype: string
  - name: founder_origin
    dtype: string
  - name: speaker_role
    dtype: string
  - name: 1a_program_subsidy
    dtype: int64
  - name: 1b_infrastructure
    dtype: int64
  - name: 1c_expectation_toward_institutions
    dtype: int64
  - name: 2a_mentors_advisors
    dtype: int64
  - name: 2b_investors_informal
    dtype: int64
  - name: 2c_peer_community
    dtype: int64
  - name: 2d_openness_access
    dtype: int64
  - name: 3a_discouragement
    dtype: int64
  - name: 3b_absence_of_support
    dtype: int64
  - name: 3c_stigma_failure_fear
    dtype: int64
  - name: 4a_emotional_support
    dtype: int64
  - name: 4b_instrumental_support
    dtype: int64
  - name: 4c_role_models_family
    dtype: int64
  - name: emergent_support_theme
    dtype: int64
  - name: none
    dtype: int64
  - name: emergent_note
    dtype: string
  - name: evidence_quote
    dtype: string
  - name: confidence
    dtype: string
  - name: _raw_response
    dtype: string
  - name: _coding_error
    dtype: string
  - name: _coding_warning
    dtype: string
  - name: run_id
    dtype: string
  - name: model
    dtype: string
  splits:
  - name: train
    num_bytes: 185454926
    num_examples: 130130
  download_size: 35449387
  dataset_size: 185454926
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---