VAGOsolutions/SauerkrautLM-ColQwen3-8b-v0.1

SauerkrautLM-ColQwen3-8b-v0.1

<p align="center"> <img src="https://vago-solutions.ai/wp-content/uploads/2025/12/Sauerkrautlm-colpali-scaled.png" alt="VAGO Solutions Logo" width="75%"/> </p>

🏆 #1 among 128-dim Models | State-of-the-Art Visual Document Retrieval

SauerkrautLM-ColQwen3-8b-v0.1 is the best-performing 128-dimensional embedding model for visual document retrieval, achieving 91.08 NDCG@5 on ViDoRe v1 - the highest score among all models with 128-dim embeddings.

<p align="center"> <img src="https://raw.githubusercontent.com/VAGOsolutions/sauerkrautlm-colpali/main/assets/benchmark128dimv1.png" alt="ViDoRe v1 Benchmark - 128-dim Models" width="100%"/> </p>

🎯 Why Visual Document Retrieval?

Traditional OCR-based retrieval loses layout, tables, and visual context. Our visual approach:

• —✅ No OCR errors - Direct visual understanding
• —✅ Layout-aware - Understands tables, forms, charts
• —✅ End-to-end - Single model, no pipeline complexity

🏆 Key Achievements

128-dim Models Comparison (XLarge Category)

Our 8B model achieves the highest ViDoRe v1 and v3 scores among ALL 128-dim models!

Detailed Benchmark Results

<details> <summary><b>📊 ViDoRe v1 (NDCG@5) - Click to expand</b></summary>

</details>

<details> <summary><b>📊 MTEB v1+v2 (NDCG@5) - Click to expand</b></summary>

ViDoRe v1 Tasks: | Task | Score | |------|-------| | ArxivQA | 93.80 🥇 | | DocVQA | 64.69 | | InfoVQA | 94.51 | | ShiftProject | 90.41 | | SyntheticDocQA-AI | 98.65 | | SyntheticDocQA-Energy | 96.52 | | SyntheticDocQA-Gov | 96.79 | | SyntheticDocQA-Health | 99.26 | | TabFQuAD | 92.18 | | TATDQA | 84.04 🥇 |

ViDoRe v2 Tasks (Multilingual): | Task | Score | |------|-------| | ViDoRe-v2-2BioMed | 63.26 | | ViDoRe-v2-2Econ | 57.98 | | ViDoRe-v2-2ESG-HL | 70.77 | | ViDoRe-v2-2ESG | 57.85 | | Combined Average | 82.91 |

</details>

<details> <summary><b>📊 ViDoRe v3 (NDCG@10) - Click to expand</b></summary>

</details>

Overall Summary (128-dim Models)

📋 Summary Tables

128-dim Models Comparison

<p align="center"> <img src="https://raw.githubusercontent.com/VAGOsolutions/sauerkrautlm-colpali/main/assets/tablesummary128dim.png" alt="128-dim Models Summary" width="100%"/> </p>

Comparison vs High-dim Models

<p align="center"> <img src="https://raw.githubusercontent.com/VAGOsolutions/sauerkrautlm-colpali/main/assets/tablesummaryhighdim_comparison.png" alt="High-dim Comparison" width="100%"/> </p>

✨ Key Features

• —🏆 #1 in 128-dim Class: Best ViDoRe v1 and v3 scores among all 128-dim models
• —⚡ Compact Embeddings: 128-dimensional (same as ColPali, 2.5x smaller than tomoro)
• —🌍 Multilingual: Trained on 6 languages (EN, DE, FR, ES, IT, PT)
• —📄 High Resolution: Supports up to 1540 visual tokens per image
• —🔧 MTEB Compatible: Standardized evaluation and easy integration
• —💻 Full Code: github.com/VAGOsolutions/sauerkrautlm-colpali

Model Details

Training

Hardware & Configuration

Datasets

Installation & Usage

Sentence Transformers

This model can be used with Sentence Transformers as a multi-vector (ColBERT-style late interaction) retriever via the MultiVectorEncoder:

pip install "sentence-transformers[image]>=6.0.0"

from sentence_transformers import MultiVectorEncoder

model = MultiVectorEncoder("VAGOsolutions/SauerkrautLM-ColQwen3-8b-v0.1")

queries = [
    "What is the variable represented on the y-axis of the graph?",
    "Total outlay is maximum in which year?",
]
images = [
    "https://huggingface.co/datasets/sentence-transformers/example-documents/resolve/main/doc1.jpg",
    "https://huggingface.co/datasets/sentence-transformers/example-documents/resolve/main/doc2.jpg",
]

query_embeddings = model.encode_query(queries)
image_embeddings = model.encode_document(images)
print(query_embeddings[0].shape, image_embeddings[0].shape)
# torch.Size([25, 128]) torch.Size([1251, 128])

# Diagonal should have higher scores
scores = model.similarity(query_embeddings, image_embeddings)
print(scores)
# tensor([[14.9121,  9.1426],
#         [ 5.8672, 15.3125]], device='cuda:0')

SauerkrautLM ColPali

⚠️ Important: Install our package first before loading the model:

pip install git+https://github.com/VAGOsolutions/sauerkrautlm-colpali

import torch
from PIL import Image
from sauerkrautlm_colpali.models import ColQwen3, ColQwen3Processor

model_name = "VAGOsolutions/SauerkrautLM-ColQwen3-8b-v0.1"

model = ColQwen3.from_pretrained(
    model_name,
    torch_dtype=torch.bfloat16,
    attn_implementation="flash_attention_2",
    device_map="cuda:0",
).eval()

processor = ColQwen3Processor.from_pretrained(model_name)

# Process inputs
images = [Image.open("document.png")]
queries = ["What is the main topic?"]

batch_images = processor.process_images(images).to(model.device)
batch_queries = processor.process_queries(queries).to(model.device)

with torch.no_grad():
    image_embeddings = model(**batch_images)
    query_embeddings = model(**batch_queries)

scores = processor.score(query_embeddings, image_embeddings)

📊 Additional Benchmark Visualizations

MTEB v1+v2 Benchmark (128-dim Models)

<p align="center"> <img src="https://raw.githubusercontent.com/VAGOsolutions/sauerkrautlm-colpali/main/assets/benchmark128dimv1v2.png" alt="MTEB v1+v2 Benchmark - 128-dim Models" width="100%"/> </p>

ViDoRe v3 Benchmark (128-dim Models)

<p align="center"> <img src="https://raw.githubusercontent.com/VAGOsolutions/sauerkrautlm-colpali/main/assets/benchmark128dimv3.png" alt="ViDoRe v3 Benchmark - 128-dim Models" width="100%"/> </p>

Our Models vs High-dim Models

<p align="center"> <img src="https://raw.githubusercontent.com/VAGOsolutions/sauerkrautlm-colpali/main/assets/benchmarkoursvshighdimv1.png" alt="ViDoRe v1 - Our Models vs High-dim" width="100%"/> </p>

Citation

@misc{sauerkrautlm-colpali-2025,
  title={SauerkrautLM-ColPali: Multi-Vector Vision Retrieval Models},
  author={David Golchinfar},
  organization={VAGO Solutions},
  year={2025},
  url={https://github.com/VAGOsolutions/sauerkrautlm-colpali}
}

Contact

• —VAGO Solutions: https://vago-solutions.ai
• —GitHub: https://github.com/VAGOsolutions
• —Email: info@vago-solutions.ai