README.md

---
dataset_info:
  features:
  - name: prompt
    dtype: string
  - name: completion
    dtype: string
  - name: states
    sequence: int64
  - name: problem_states
    sequence: int64
  - name: contexts
    sequence: int64
  - name: actions
    sequence: int64
  - name: text
    dtype: string
  splits:
  - name: train
    num_bytes: 118731236
    num_examples: 200000
  - name: train_rl
    num_bytes: 76512
    num_examples: 128
  - name: test
    num_bytes: 305080
    num_examples: 512
  download_size: 32291720
  dataset_size: 119112828
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: train_rl
    path: data/train_rl-*
  - split: test
    path: data/test-*
---