README.md

---
dataset_info:
  features:
  - name: latent
    dtype: int64
  - name: count
    dtype: float64
  - name: total
    dtype: float64
  - name: mean
    dtype: float64
  - name: layer_mean
    dtype: float64
  - name: layer_var
    dtype: float64
  - name: layer_std
    dtype: float64
  - name: counts
    sequence: float64
  - name: totals
    sequence: float64
  - name: means
    sequence: float64
  - name: probs
    sequence: float64
  splits:
  - name: train
    num_bytes: 22413312
    num_examples: 49152
  download_size: 10317298
  dataset_size: 22413312
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---