---
dataset_info:
  features:
  - name: task_id
    dtype: int64
  - name: repository
    dtype: string
  - name: commit_hash
    dtype: string
  - name: filepath
    dtype: string
  - name: original_method
    dtype: string
  - name: context
    list:
    - name: filepath
      dtype: string
    - name: code
      dtype: string
  - name: masking
    struct:
    - name: fim_idx
      dtype: int64
    - name: mask_type
      dtype: string
    - name: n_masked_tokens
      dtype: int64
    - name: prefix
      dtype: string
    - name: suffix
      dtype: string
    - name: completion
      dtype: string
  splits:
  - name: train
    num_bytes: 17456040
    num_examples: 2264
  download_size: 579465
  dataset_size: 17456040
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---