Dataset Viewer
Auto-converted to Parquet Duplicate
text_id
int64
0
1.03k
chunk_id
int64
0
0
pixel_values
listlengths
16
16
grapheme_token_ids
listlengths
3
19
llama_token_ids
listlengths
3
15
text
stringlengths
2
30
0
0
[[255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,25(...TRUNCATED)
[ 1, 19, 0, 2 ]
[ 1, 3500, 273, 2 ]
papan
1
0
[[255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,25(...TRUNCATED)
[ 1, 19, 116, 2 ]
[ 1, 10203, 29875, 2 ]
pagi
2
0
[[255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,25(...TRUNCATED)
[ 1, 110, 110, 22, 2 ]
[ 1, 367, 16863, 2 ]
bebek
3
0
[[255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,25(...TRUNCATED)
[ 1, 0, 50, 41, 2 ]
[ 1, 286, 5893, 29875, 2 ]
menteri
4
0
[[255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,25(...TRUNCATED)
[ 1, 156, 282, 2 ]
[ 1, 289, 574, 29895, 574, 2 ]
bangkang
5
0
[[255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,25(...TRUNCATED)
[ 1, 194, 194, 2 ]
[ 1, 2594, 1646, 2 ]
barbar
6
0
[[255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,25(...TRUNCATED)
[ 1, 6, 0, 2 ]
[ 1, 472, 1450, 2 ]
ataw
7
0
[[255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,25(...TRUNCATED)
[ 1, 103, 135, 2 ]
[ 1, 282, 686, 29884, 2 ]
pungu
8
0
[[255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,25(...TRUNCATED)
[ 1, 66, 208, 2 ]
[ 1, 21062, 272, 2 ]
kotor
9
0
[[255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,25(...TRUNCATED)
[ 1, 13, 0, 2 ]
[ 1, 15296, 1794, 2 ]
sabai
End of preview. Expand in Data Studio

Lampung PixelGPT Dataset

This dataset contains preprocessed Lampung text data for training PixelGPT models.

Dataset Statistics

  • Language: Lampung (lampung)
  • Total samples: 1,029
  • Train samples: 945
  • Test samples: 84

Tokenizers

  • Grapheme tokenizer: izzako/sunda-llama-tokenizer
  • LLaMA tokenizer: ernie-research/DualGPT

Features

  • text_id: Document identifier
  • chunk_id: Chunk identifier within document
  • pixel_values: Rendered pixel representation of aksara text
  • grapheme_token_ids: Token IDs from grapheme-based tokenizer (with EOS token)
  • llama_token_ids: Token IDs from LLaMA tokenizer (with EOS token)
  • text: Original text chunk

Renderer Configuration

  • Renderer path: ../../renderers/lampung_renderer

Usage

from datasets import load_dataset

# Load the dataset
dataset = load_dataset("izzako/lampung-pixelgpt")

# Access train and test splits
train_data = dataset['train']
test_data = dataset['test']

# Example: Get first sample
sample = train_data[0]
print(sample['text'])

Citation

If you use this dataset, please cite:

@dataset{lampung_pixelgpt,
  title = {Lampung PixelGPT Dataset},
  author = {Musa Izzanardi Wijanarko},
  year = {2025},
  publisher = {Hugging Face},
  url = {https://huggingface.co/datasets/izzako/lampung-pixelgpt}
}
Downloads last month
30