Stack_TokLlama / README.md
Trevor Dohm
Tokenized bro
d1a261c verified
|
raw
history blame
329 Bytes
metadata
dataset_info:
  features:
    - name: input_ids
      sequence: int32
    - name: attention_mask
      sequence: int8
  splits:
    - name: train
      num_bytes: 7827160
      num_examples: 1100
  download_size: 1628463
  dataset_size: 7827160
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/bro/train-*