---
language:
- en
license: apache-2.0
tags:
- colbert
- embeddings
- information-retrieval
- ms-marco
- answer-dot-ai
- pylate
task_categories:
- feature-extraction
- text-retrieval
size_categories:
- 1M<n<10M
pretty_name: MS MARCO ColBERT Embeddings
configs:
- config_name: corpus
  data_files:
  - split: train
    path: data/corpus/*.parquet
- config_name: queries
  data_files:
  - split: train
    path: data/queries/*.parquet
- config_name: qrels
  data_files:
  - split: train
    path: data/qrels/train.parquet
---

# MS MARCO ColBERT Embeddings

Pre-computed ColBERT embeddings for [MS MARCO](https://huggingface.co/datasets/BeIR/msmarco) using [PyLate](https://github.com/lightonai/pylate) and [answerdotai/answerai-colbert-small-v1](https://huggingface.co/answerdotai/answerai-colbert-small-v1).

## Dataset Structure

The dataset contains:
- `data/corpus/`: 177 parquet files with document embeddings
- `data/queries/`: 11 parquet files with query embeddings
- `data/qrels/train.parquet`: Relevance judgments (532,751 pairs)

## Usage

```python
from datasets import load_dataset

# Load from directory (recommended for large datasets)
corpus = load_dataset("parquet", data_dir="hf://datasets/WenxingZhu/msmarco_answerai_colbert_small_embeddings/data/corpus")
queries = load_dataset("parquet", data_dir="hf://datasets/WenxingZhu/msmarco_answerai_colbert_small_embeddings/data/queries")
qrels = load_dataset("parquet", data_files="hf://datasets/WenxingZhu/msmarco_answerai_colbert_small_embeddings/data/qrels/train.parquet")
```

## Model
Generated with [answerdotai/answerai-colbert-small-v1](https://huggingface.co/answerdotai/answerai-colbert-small-v1)