[Model card] [ViDoRe Leaderboard] [Demo] [Blog Post]
Tip
For production usage in your RAG pipelines, we recommend using the byaldi package, which is a lightweight wrapper around the colpali-engine package developed by the author of the popular RAGatouille repostiory. 🐭
This repository contains the code used for training the vision retrievers in the ColPali: Efficient Document Retrieval with Vision Language Models paper. In particular, it contains the code for training the ColPali model, which is a vision retriever based on the ColBERT architecture and the PaliGemma model.
With our new model ColPali, we propose to leverage VLMs to construct efficient multi-vector embeddings in the visual space for document retrieval. By feeding the ViT output patches from PaliGemma-3B to a linear projection, we create a multi-vector representation of documents. We train the model to maximize the similarity between these document embeddings and the query embeddings, following the ColBERT method.
Using ColPali removes the need for potentially complex and brittle layout recognition and OCR pipelines with a single model that can take into account both the textual and visual content (layout, charts, ...) of a document.
| Model | Score on ViDoRe 🏆 | License | Comments | Currently supported |
|---|---|---|---|---|
| vidore/colpali | 81.3 | Gemma | • Based on google/paligemma-3b-mix-448.• Checkpoint used in the ColPali paper. |
❌ |
| vidore/colpali-v1.1 | 81.5 | Gemma | • Based on google/paligemma-3b-mix-448. |
✅ |
| vidore/colpali-v1.2 | 83.1 | Gemma | • Based on google/paligemma-3b-mix-448. |
✅ |
| vidore/colqwen2-v0.1 | 86.6 | Apache 2.0 | • Based on Qwen/Qwen2-VL-2B-Instruct.• Supports dynamic resolution. • Trained using 768 image patches per page. |
✅ |
We used Python 3.11.6 and PyTorch 2.2.2 to train and test our models, but the codebase is compatible with Python >=3.9 and recent PyTorch versions. To install the package, run:
pip install colpali-engineWarning
For ColPali versions above v1.0, make sure to install the colpali-engine package from source or with a version above v0.2.0.
import torch
from PIL import Image
from colpali_engine.models import ColPali, ColPaliProcessor
model_name = "vidore/colpali-v1.2"
model = ColPali.from_pretrained(
model_name,
torch_dtype=torch.bfloat16,
device_map="cuda:0", # or "mps" if on Apple Silicon
).eval()
processor = ColPaliProcessor.from_pretrained(model_name)
# Your inputs
images = [
Image.new("RGB", (32, 32), color="white"),
Image.new("RGB", (16, 16), color="black"),
]
queries = [
"Is attention really all you need?",
"Are Benjamin, Antoine, Merve, and Jo best friends?",
]
# Process the inputs
batch_images = processor.process_images(images).to(model.device)
batch_queries = processor.process_queries(queries).to(model.device)
# Forward pass
with torch.no_grad():
image_embeddings = model(**batch_images)
query_embeddings = model(**batch_queries)
scores = processor.score_multi_vector(query_embeddings, image_embeddings)You can find an example here. If you need an indexing system, we recommend using byaldi - RAGatouille's little sister 🐭 - which share a similar API and leverages our colpali-engine package.
To benchmark ColPali to reproduce the results on the ViDoRe leaderboard, it is recommended to use the vidore-benchmark package.
By superimposing the late interaction similarity maps on top of the original image, we can visualize the most salient image patches with respect to each term of the query, yielding interpretable insights into model focus zones.
To use the interpretability module, you need to install the colpali-engine[interpretability] package:
pip install colpali-engine[interpretability]Then, after generating your embeddings with ColPali, use the following code to plot the similarity maps for each query token:
import torch
from PIL import Image
from colpali_engine.interpretability import (
get_similarity_maps_from_embeddings,
plot_all_similarity_maps,
)
from colpali_engine.models import ColPali, ColPaliProcessor
from colpali_engine.utils.torch_utils import get_torch_device
model_name = "vidore/colpali-v1.2"
device = get_torch_device("auto")
# Load the model
model = ColPali.from_pretrained(
model_name,
torch_dtype=torch.bfloat16,
device_map=device,
).eval()
# Load the processor
processor = ColPaliProcessor.from_pretrained(model_name)
# Load the image and query
image = Image.open("shift_kazakhstan.jpg")
query = "Quelle partie de la production pétrolière du Kazakhstan provient de champs en mer ?"
# Preprocess inputs
batch_images = processor.process_images([image]).to(device)
batch_queries = processor.process_queries([query]).to(device)
# Forward passes
with torch.no_grad():
image_embeddings = model.forward(**batch_images)
query_embeddings = model.forward(**batch_queries)
# Get the number of image patches
n_patches = processor.get_n_patches(image_size=image.size, patch_size=model.patch_size)
# Get the tensor mask to filter out the embeddings that are not related to the image
image_mask = processor.get_image_mask(batch_images)
# Generate the similarity maps
batched_similarity_maps = get_similarity_maps_from_embeddings(
image_embeddings=image_embeddings,
query_embeddings=query_embeddings,
n_patches=n_patches,
image_mask=image_mask,
)
# Get the similarity map for our (only) input image
similarity_maps = batched_similarity_maps[0] # (query_length, n_patches_x, n_patches_y)
# Tokenize the query
query_tokens = processor.tokenizer.tokenize(query)
# Plot and save the similarity maps for each query token
plots = plot_all_similarity_maps(
image=image,
query_tokens=query_tokens,
similarity_maps=similarity_maps,
)
for idx, (fig, ax) in enumerate(plots):
fig.savefig(f"similarity_map_{idx}.png")For a more detailed example, you can refer to the interpretability notebooks from the ColPali Cookbooks 👨🏻🍳 repository.
To keep a lightweight repository, only the essential packages were installed. In particular, you must specify the dependencies to use the training script for ColPali. You can do this using the following command:
pip install "colpali-engine[train]"All the model configs used can be found in scripts/configs/ and rely on the configue package for straightforward configuration. They should be used with the train_colbert.py script.
USE_LOCAL_DATASET=0 python scripts/train/train_colbert.py scripts/configs/pali/train_colpali_docmatix_hardneg_model.yamlor using accelerate:
accelerate launch scripts/train/train_colbert.py scripts/configs/pali/train_colpali_docmatix_hardneg_model.yamlsbatch --nodes=1 --cpus-per-task=16 --mem-per-cpu=32GB --time=20:00:00 --gres=gpu:1 -p gpua100 --job-name=colidefics --output=colidefics.out --error=colidefics.err --wrap="accelerate launch scripts/train/train_colbert.py scripts/configs/pali/train_colpali_docmatix_hardneg_model.yaml"
sbatch --nodes=1 --time=5:00:00 -A cad15443 --gres=gpu:8 --constraint=MI250 --job-name=colpali --wrap="python scripts/train/train_colbert.py scripts/configs/pali/train_colpali_docmatix_hardneg_model.yaml"To reproduce the results from the paper, you should checkout to the v0.1.1 tag or install the corresponding colpali-engine package release using:
pip install colpali-engine==0.1.1ColPali: Efficient Document Retrieval with Vision Language Models
Authors: Manuel Faysse*, Hugues Sibille*, Tony Wu*, Bilel Omrani, Gautier Viaud, Céline Hudelot, Pierre Colombo (* denotes equal contribution)
@misc{faysse2024colpaliefficientdocumentretrieval,
title={ColPali: Efficient Document Retrieval with Vision Language Models},
author={Manuel Faysse and Hugues Sibille and Tony Wu and Bilel Omrani and Gautier Viaud and Céline Hudelot and Pierre Colombo},
year={2024},
eprint={2407.01449},
archivePrefix={arXiv},
primaryClass={cs.IR},
url={https://arxiv.org/abs/2407.01449},
}