README.md

---
dataset_info:
  features:
  - name: tgt
    dtype: string
  - name: preds_greedy
    dtype: string
  - name: preds_beam10
    dtype: string
  - name: preds_topk20
    dtype: string
  - name: edit_dist_greedy
    dtype: int64
  - name: edit_dist_beam10
    dtype: int64
  - name: edit_dist_topk20
    dtype: int64
  splits:
  - name: train
    num_bytes: 2338859
    num_examples: 4096
  download_size: 1263302
  dataset_size: 2338859
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---