Fast Multimodal Semantic Deduplication & Filtering

Quickstart • Main Features • Usage • Benchmarks

SemHash is a lightweight, multimodal library for semantic deduplication, outlier filtering, and representative sample selection. Text works out of the box with fast Model2Vec embeddings, and images, audio, and other modalities are supported with custom encoders.

SemHash supports both single-dataset operations (clean a training set) and cross-dataset operations (deduplicate test against train). It works with simple lists and complex multi-column datasets, and includes inspection tools to help you understand and refine results. All operations use Vicinity for efficient similarity search.

Quickstart

Install the package with:

pip install semhash

Text Deduplication, Filtering & Representative Sampling

Deduplicate a single dataset, filter outliers, and find representative samples with the following code (note: the examples assume you have datasets installed, which you can install with pip install datasets):

from datasets import load_dataset
from semhash import SemHash

# Load a dataset to deduplicate
texts = load_dataset("ag_news", split="train")["text"]

# Initialize a SemHash instance
semhash = SemHash.from_records(records=texts)

# Deduplicate the texts
deduplicated_texts = semhash.self_deduplicate().selected

# Filter outliers
filtered_texts = semhash.self_filter_outliers().selected

# Find representative texts
representative_texts = semhash.self_find_representative().selected

Image Deduplication, Filtering & Representative Sampling

Deduplicate an image dataset, filter outliers, and find representative samples using a vision model (requires pip install sentence-transformers):

from datasets import load_dataset
from sentence_transformers import SentenceTransformer
from semhash import SemHash

# Load an image dataset and vision model
model = SentenceTransformer('clip-ViT-B-32')
dataset = load_dataset("uoft-cs/cifar10", split="test")

# Initialize a SemHash instance with the 'img' column
semhash = SemHash.from_records(list(dataset), columns=["img"], model=model)

# Deduplicate the images
deduplicated_images = semhash.self_deduplicate().selected

# Filter outliers
filtered_images = semhash.self_filter_outliers().selected

# Find representative images
representative_images = semhash.self_find_representative().selected

Cross-Dataset Deduplication, Filtering & Representative Sampling

Deduplicate across two datasets, filter outliers, and find representative samples (e.g., eliminating train/test leakage):

from datasets import load_dataset
from semhash import SemHash

# Load two datasets to deduplicate
train_texts = load_dataset("ag_news", split="train")["text"]
test_texts = load_dataset("ag_news", split="test")["text"]

# Initialize a SemHash instance with the training data
semhash = SemHash.from_records(records=train_texts)

# Deduplicate the test data against the training data, optionally with a specific threshold
deduplicated_test_texts = semhash.deduplicate(records=test_texts, threshold=0.9).selected

# Filter outliers from the test data against the training data,
# optionally with a specific percentage
filtered_test_texts = semhash.filter_outliers(records=test_texts, outlier_percentage=0.1).selected

# Find representative texts in the test data against the training data,
# optionally with a specific selection size
representative_test_texts = semhash.find_representative(records=test_texts, selection_size=10).selected

Multi-Column Deduplication

Deduplicate multi-column datasets (e.g., deduplicating a QA dataset):

from datasets import load_dataset
from semhash import SemHash

# Load the dataset
dataset = load_dataset("squad_v2", split="train")

# Convert the dataset to a list of dictionaries
records = [dict(row) for row in dataset]

# Initialize SemHash with the columns to deduplicate
semhash = SemHash.from_records(records=records, columns=["question", "context"])

# Deduplicate the records
deduplicated_records = semhash.self_deduplicate().selected

The deduplicate and self_deduplicate functions return a DeduplicationResult. This object stores the deduplicated corpus, a set of duplicate objects (along with the objects that caused duplication), and several useful functions to further inspect the deduplication result.

The filter_outliers, self_filter_outliers, find_representative, and self_find_representative functions return a FilterResult. This object stores the found outliers/representative samples.

For both the DeduplicationResult and FilterResult objects, you can easily view the filtered records with the selected attribute (e.g. to view outliers: outliers = semhash.self_filter_outliers().filtered)

Inspecting Deduplication Results

The DeduplicationResult object provides powerful tools for understanding and refining your deduplication:

from datasets import load_dataset
from semhash import SemHash

# Load and deduplicate a dataset
texts = load_dataset("ag_news", split="train")["text"]
semhash = SemHash.from_records(records=texts)
result = semhash.self_deduplicate()

# Access deduplicated and duplicate records
deduplicated_texts = result.selected
duplicate_texts = result.filtered

# View deduplication statistics
print(f"Duplicate ratio: {result.duplicate_ratio}")
print(f"Exact duplicate ratio: {result.exact_duplicate_ratio}")

# Find edge cases to tune your threshold
least_similar = result.get_least_similar_from_duplicates(n=5)

# Adjust threshold without re-deduplicating
result.rethreshold(0.95)

# View each kept record with its duplicate cluster
for item in result.selected_with_duplicates:
    print(f"Kept: {item.record}")
    print(f"Duplicates: {item.duplicates}")  # List of (duplicate_text, similarity_score)

Main Features

• Fast: SemHash uses model2vec to embed texts and vicinity to perform similarity search, making it extremely fast.
• Scalable: SemHash can deduplicate & filter large datasets with millions of records thanks to the ANN backends in Vicinity.
• Flexible: SemHash can be used to deduplicate & filter a single dataset or across two datasets, and can also be used to deduplicate & filter multi-column datasets (such as QA datasets).
• Lightweight: SemHash is a lightweight package with minimal dependencies, making it easy to install and use.
• Explainable: Easily inspect the duplicates and what caused them with the DeduplicationResult object. You can also view the lowest similarity duplicates to find the right threshold for deduplication for your dataset.

Usage

The following examples show the various ways you can use SemHash to deduplicate datasets, filter outliers, and find representative samples. These examples assume you have the datasets library installed, which you can install with pip install datasets.

Deduplicate, filter outliers, and find representative samples on a single dataset

The following code snippet shows how to deduplicate a single dataset, filter outliers, and find representative samples using SemHash (in this example, the train split of the AG News dataset):

from datasets import load_dataset
from semhash import SemHash

# Load a dataset to deduplicate
texts = load_dataset("ag_news", split="train")["text"]

# Initialize a SemHash instance
semhash = SemHash.from_records(records=texts)

# Deduplicate the texts
deduplicated_texts = semhash.self_deduplicate().selected

# Filter outliers
filtered_texts = semhash.self_filter_outliers().selected

# Find representative texts
representative_texts = semhash.self_find_representative().selected

Deduplicate, filter outliers, and find representative samples across two datasets

The following code snippet shows how to deduplicate across two datasets, filter outliers, and find representative samples using SemHash (in this example, the train/test split of the AG News dataset):

from datasets import load_dataset
from semhash import SemHash

# Load two datasets to deduplicate
train_texts = load_dataset("ag_news", split="train")["text"]
test_texts = load_dataset("ag_news", split="test")["text"]

# Initialize a SemHash instance with the training data
semhash = SemHash.from_records(records=train_texts)

# Deduplicate the test data against the training data
deduplicated_test_texts = semhash.deduplicate(records=test_texts).selected

# Filter outliers from the test data
filtered_test_texts = semhash.filter_outliers(records=test_texts).selected

# Find representative texts in the test data
representative_test_texts = semhash.find_representative(records=test_texts).selected

Deduplicate, filter outliers, and find representative samples on multi-column datasets

The following code snippet shows how to deduplicate multi-column datasets, filter outliers, and find representative samples using SemHash (in this example, the train split of the QA dataset SQuAD 2.0, which consists of questions, contexts, and answers):

from datasets import load_dataset
from semhash import SemHash

# Load the dataset
dataset = load_dataset("squad_v2", split="train")

# Convert the dataset to a list of dictionaries
records = [dict(row) for row in dataset]

# Initialize SemHash with the columns to deduplicate
semhash = SemHash.from_records(records=records, columns=["question", "context"])

# Deduplicate the records
deduplicated_records = semhash.self_deduplicate().selected

# Filter outliers from the records
filtered_records = semhash.self_filter_outliers().selected

# Find representative samples in the records
representative_records = semhash.self_find_representative().selected

Deduplicate, filter outliers, and find representative samples on image datasets

You can bring your own encoder for any modality by implementing the Encoder protocol. Here's an example using a vision model from timm for image deduplication:

from datasets import load_dataset
import timm
import torch
from semhash import SemHash

# Requires: pip install timm torch datasets

# Create a custom image encoder
class VisionEncoder:
    """Custom encoder using timm models. Implements the Encoder protocol."""

    def __init__(self, model_name: str = "mobilenetv3_small_100.lamb_in1k"):
        self.model = timm.create_model(model_name, pretrained=True, num_classes=0).eval()
        data_config = timm.data.resolve_model_data_config(self.model)
        self.transform = timm.data.create_transform(**data_config, is_training=False)

    def encode(self, inputs, batch_size: int = 128):
        """Encode a batch of PIL images into embeddings."""
        import numpy as np

        # Convert grayscale to RGB if needed
        rgb_inputs = [img.convert("RGB") if img.mode != "RGB" else img for img in inputs]

        # Process in batches to avoid memory issues
        all_embeddings = []
        with torch.no_grad():
            for i in range(0, len(rgb_inputs), batch_size):
                batch_inputs = rgb_inputs[i : i + batch_size]
                batch = torch.stack([self.transform(img) for img in batch_inputs])
                embeddings = self.model(batch).numpy()
                all_embeddings.append(embeddings)

        return np.vstack(all_embeddings)

# Load image dataset
dataset = load_dataset("uoft-cs/cifar10", split="test")
train_data = [{"img": img, "id": i} for i, img in enumerate(dataset["img"][:100])]
test_data = [{"img": img, "id": i} for i, img in enumerate(dataset["img"][100:150])]

# Initialize SemHash with the custom vision encoder
semhash = SemHash.from_records(train_data, columns=["img"], model=VisionEncoder())

# Single-dataset operations
deduplicated = semhash.self_deduplicate().selected
outliers = semhash.self_filter_outliers().selected
representatives = semhash.self_find_representative().selected

# Cross-dataset operations
test_deduplicated = semhash.deduplicate(test_data).selected
test_outliers = semhash.filter_outliers(test_data).selected
test_representatives = semhash.find_representative(test_data, selection_size=10).selected

The Encoder protocol requires only an encode(inputs, **kwargs) method that returns a numpy array. This makes it easy to integrate any embedding model for any modality.

Using custom encoders

The following code snippet shows how to use a custom encoder with SemHash:

from datasets import load_dataset
from model2vec import StaticModel
from semhash import SemHash

# Load a dataset to deduplicate
texts = load_dataset("ag_news", split="train")["text"]

# Load an embedding model (in this example, a multilingual model)
model = StaticModel.from_pretrained("minishlab/potion-multilingual-128M")

# Initialize a SemHash with the model and custom encoder
semhash = SemHash.from_records(records=texts, model=model)

# Deduplicate the texts
deduplicated_texts = semhash.self_deduplicate()

Any encoder can be used that adheres to our encoder protocol. For example, any sentence-transformers model can be used as an encoder:

from datasets import load_dataset
from semhash import SemHash
from sentence_transformers import SentenceTransformer

# Load a dataset to deduplicate
texts = load_dataset("ag_news", split="train")["text"]

# Load a sentence-transformers model
model = SentenceTransformer("sentence-transformers/all-MiniLM-L6-v2")

# Initialize a SemHash with the model and custom encoder
semhash = SemHash.from_records(records=texts, model=model)

# Deduplicate the texts
deduplicated_texts = semhash.self_deduplicate()

Using custom ANN backends

By default, we use USearch as the ANN (approximate-nearest neighbors) backend for deduplication. We recommend keeping this since the recall for smaller datasets is ~100%, and it's needed for larger datasets (>1M samples) since these will take too long to deduplicate without ANN. If you want to use a flat/exact-matching backend, you can set ann_backend=Backend.BASIC in the SemHash constructor:

from semhash import SemHash
from vicinity import Backend

semhash = SemHash.from_records(records=texts, ann_backend=Backend.BASIC)

Any backend from Vicinity can be used with SemHash. The following code snippet shows how to use FAISS with a custom nlist parameter:

from datasets import load_dataset
from semhash import SemHash
from vicinity import Backend

semhash = SemHash.from_records(records=texts, ann_backend=Backend.FAISS, nlist=50)

For the full list of supported ANN backends and args, see the Vicinity docs.

Using Pandas DataFrames

You can easily use Pandas DataFrames with SemHash. The following code snippet shows how to deduplicate a Pandas DataFrame:

import pandas as pd
from datasets import load_dataset
from semhash import SemHash

# Load a dataset as a pandas dataframe
dataframe = load_dataset("ag_news", split="train").to_pandas()

# Convert the dataframe to a list of dictionaries
dataframe = dataframe.to_dict(orient="records")

# Initialize a SemHash instance with the columns to deduplicate
semhash = SemHash.from_records(records=dataframe, columns=["text"])

# Deduplicate the texts
deduplicated_records = semhash.self_deduplicate().selected

# Convert the deduplicated records back to a pandas dataframe
deduplicated_dataframe = pd.DataFrame(deduplicated_records)

Initializing from embeddings

You can also initialize SemHash from pre-computed embeddings. The following code snippet shows how to do this:

from datasets import load_dataset
from model2vec import StaticModel
from semhash import SemHash

# Load a dataset
texts = load_dataset("ag_news", split="train")["text"]

# Load an embedding model
model = StaticModel.from_pretrained("minishlab/potion-base-8M")

# Create embeddings
embeddings = model.encode(texts)

# Initialize SemHash from embeddings
semhash = SemHash.from_embeddings(embeddings=embeddings, records=texts, model=model)

# Deduplicate, filter outliers, and find representative samples
deduplicated_texts = semhash.self_deduplicate().selected
filtered_texts = semhash.self_filter_outliers().selected
representative_texts = semhash.self_find_representative().selected

Initializing from a HuggingFace Dataset

You can easily use SemHash with HuggingFace Datasets by converting them to a list:

from datasets import load_dataset
from semhash import SemHash

# Load a HuggingFace dataset
dataset = load_dataset("ag_news", split="train")

# Convert to list and initialize SemHash
semhash = SemHash.from_records(records=list(dataset), columns=["text"])

# Deduplicate, filter outliers, and find representative samples
deduplicated_texts = semhash.self_deduplicate().selected
filtered_texts = semhash.self_filter_outliers().selected
representative_texts = semhash.self_find_representative().selected

This also works with multi-column datasets:

from datasets import load_dataset
from semhash import SemHash

# Load a multi-column dataset
dataset = load_dataset("squad_v2", split="train")

# Convert to list and initialize with multiple columns
semhash = SemHash.from_records(records=list(dataset), columns=["question", "context"])

# Deduplicate the records
deduplicated_records = semhash.self_deduplicate().selected

Benchmarks

SemHash is extremely fast and scales to large datasets with millions of records. We've benchmarked both text and image deduplication across a variety of datasets. For example, deduplicating text 1.8M records takes only ~83 seconds on CPU.

For detailed benchmark results and analysis, see the benchmarks directory.

Running Benchmarks

# Run text benchmarks
make benchmark-text

# Run image benchmarks
make benchmark-image

# Run all benchmarks
make benchmark

License

MIT

Citing

If you use SemHash in your research, please cite the following:

@software{minishlab2025semhash,
  author       = {{van Dongen}, Thomas and Stephan Tulkens},
  title        = {SemHash: Fast Multimodal Semantic Deduplication \& Filtering},
  year         = {2025},
  publisher    = {Zenodo},
  doi          = {10.5281/zenodo.17265942},
  url          = {https://github.com/MinishLab/semhash},
  license      = {MIT}
}