README.md

---
dataset_info:
  features:
  - name: task_id
    dtype: string
  - name: trial_number
    dtype: int32
  - name: task_description
    dtype: string
  - name: start_url
    dtype: string
  - name: is_success
    dtype: bool
  - name: is_shortest
    dtype: bool
  - name: evaluator_thoughts
    dtype: string
  - name: evaluator_status
    dtype: string
  - name: run_error
    dtype: string
  - name: step_index
    dtype: int32
  - name: url_at_step
    dtype: string
  - name: action
    dtype: string
  - name: screenshot
    dtype: image
  splits:
  - name: train
    num_bytes: 6584752.0
    num_examples: 30
  download_size: 6404968
  dataset_size: 6584752.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---