README.md

metadata

license: cc-by-sa-4.0
task_categories:
  - text-retrieval
  - sentence-similarity
language:
  - en
tags:
  - wikipedia
  - embeddings
  - splade
  - bge-m3
  - dense-retrieval
  - sparse-retrieval
  - hybrid-search
size_categories:
  - 1M<n<10M
configs:
  - config_name: default
    data_files:
      - split: train
        path: train/*.parquet
      - split: test
        path: test/*.parquet

Wikipedia English with SPLADE and BGE-M3

Pre-computed SPLADE sparse and BGE-M3 dense embeddings for 6.4M English Wikipedia articles.

Dataset Description

• Source: HuggingFaceFW/clean-wikipedia (English)
• Size:
  • Train: 6,406,711 documents
  • Test: 256 queries
• Format:
  • Train: 7 Parquet files (~1M records each) in train/ directory
  • Test: 1 Parquet file in test/ directory
• Embeddings:
  • Sparse: SPLADE PP - ~265 non-zero dims
  • Dense: BGE-M3 - 1024 dimensions

Direct Usage

HuggingFace automatically discovers parquet files. You can load this dataset directly:

from datasets import load_dataset

# Stream the entire dataset (recommended for large dataset)
dataset = load_dataset("Sicheng-Chroma/wikipedia-en-splade-bge", streaming=True)

# Load specific splits
train_dataset = load_dataset("Sicheng-Chroma/wikipedia-en-splade-bge", split="train")
test_dataset = load_dataset("Sicheng-Chroma/wikipedia-en-splade-bge", split="test")

# Load specific chunk
dataset = load_dataset("Sicheng-Chroma/wikipedia-en-splade-bge", 
                       data_files="train/train-00000-of-00007.parquet")

# Load multiple chunks
dataset = load_dataset("Sicheng-Chroma/wikipedia-en-splade-bge", 
                       data_files=["train/train-00000-of-00007.parquet", "train/train-00001-of-00007.parquet"])

Dataset Structure

Each row contains:

Field	Type	Description
text	string	Full Wikipedia article text
title	string	Article title
url	string	Wikipedia URL
sparse_embedding_indices	list[int32]	SPLADE indices (non-zero positions)
sparse_embedding_values	list[float32]	SPLADE values (weights)
dense_embedding	list[float32]	BGE-M3 1024-dim dense vector

Example Usage

from datasets import load_dataset

# Load the dataset (streaming recommended for large dataset)
ds = load_dataset("Sicheng-Chroma/wikipedia-en-splade-bge", streaming=True)

# Iterate through articles
for article in ds['train']:
    print(f"Title: {article['title']}")
    print(f"URL: {article['url']}")
    print(f"Text preview: {article['text'][:200]}...")
    print(f"Sparse embedding: {len(article['sparse_embedding_indices'])} non-zero dims")
    print(f"Dense embedding: {len(article['dense_embedding'])} dims")
    break

# Load into memory (for smaller chunks)
dataset = load_dataset("Sicheng-Chroma/wikipedia-en-splade-bge", 
                       data_files="train/train-00000-of-00007.parquet")
print(f"Loaded {len(dataset['train'])} articles")

File Organization

.
├── train/
│   ├── train-00000-of-00007.parquet - First 1M articles
│   ├── train-00001-of-00007.parquet - Next 1M articles
│   ├── train-00002-of-00007.parquet
│   ├── train-00003-of-00007.parquet
│   ├── train-00004-of-00007.parquet
│   ├── train-00005-of-00007.parquet
│   └── train-00006-of-00007.parquet - Final ~400K articles
└── test/
    └── test-00000-of-00001.parquet - 256 test queries

License

CC BY-SA 4.0 (inherited from Wikipedia)