Camel-Doc-OCR-062825

The Camel-Doc-OCR-062825 model is a fine-tuned version of Qwen2.5-VL-7B-Instruct, optimized for Document Retrieval, Content Extraction, and Analysis Recognition. Built on top of the Qwen2.5-VL architecture, this model enhances document comprehension capabilities with focused training on the Opendoc2-Analysis-Recognition dataset for superior document analysis and information extraction tasks.

Key Enhancements

• Context-Aware Multimodal Extraction and Linking for Documents: Advanced capability for understanding document context and establishing connections between multimodal elements within documents.

• Enhanced Document Retrieval: Designed to efficiently locate and extract relevant information from complex document structures and layouts.

• Superior Content Extraction: Optimized for precise extraction of structured and unstructured content from diverse document formats.

• Analysis Recognition: Specialized in recognizing and interpreting analytical content, charts, tables, and visual data representations.

• State-of-the-Art Performance Across Resolutions: Achieves competitive results on OCR and visual QA benchmarks such as DocVQA, MathVista, RealWorldQA, and MTVQA.

• Video Understanding up to 20+ minutes: Supports detailed comprehension of long-duration videos for content summarization, Q&A, and multi-modal reasoning.

• Visually-Grounded Device Interaction: Enables mobile/robotic device operation via visual inputs and text-based instructions using contextual understanding and decision-making logic.

Quick Start with Transformers

from transformers import Qwen2_5_VLForConditionalGeneration, AutoTokenizer, AutoProcessor
from qwen_vl_utils import process_vision_info

model = Qwen2_5_VLForConditionalGeneration.from_pretrained(
    "prithivMLmods/Camel-Doc-OCR-062825", torch_dtype="auto", device_map="auto"
)

processor = AutoProcessor.from_pretrained("prithivMLmods/Camel-Doc-OCR-062825")

messages = [
    {
        "role": "user",
        "content": [
            {
                "type": "image",
                "image": "https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-VL/assets/demo.jpeg",
            },
            {"type": "text", "text": "Describe this image."},
        ],
    }
]

text = processor.apply_chat_template(
    messages, tokenize=False, add_generation_prompt=True
)
image_inputs, video_inputs = process_vision_info(messages)
inputs = processor(
    text=[text],
    images=image_inputs,
    videos=video_inputs,
    padding=True,
    return_tensors="pt",
)
inputs = inputs.to("cuda")

generated_ids = model.generate(**inputs, max_new_tokens=128)
generated_ids_trimmed = [
    out_ids[len(in_ids):] for in_ids, out_ids in zip(inputs.input_ids, generated_ids)
]
output_text = processor.batch_decode(
    generated_ids_trimmed, skip_special_tokens=True, clean_up_tokenization_spaces=False
)
print(output_text)

Intended Use

This model is intended for:

• Context-aware multimodal extraction and linking for complex document structures.
• High-fidelity document retrieval and content extraction from various document formats.
• Analysis recognition of charts, graphs, tables, and visual data representations.
• Document-based question answering for educational and enterprise applications.
• Extraction and LaTeX formatting of mathematical expressions from printed or handwritten content.
• Retrieval and summarization from long documents, slides, and multi-modal inputs.
• Multilingual document analysis and structured content extraction for global use cases.
• Robotic or mobile automation with vision-guided contextual interaction.

Limitations

• May show degraded performance on extremely low-quality or occluded images.
• Not optimized for real-time applications on low-resource or edge devices due to computational demands.
• Variable accuracy on uncommon or low-resource languages/scripts.
• Long video processing may require substantial memory and is not optimized for streaming applications.
• Visual token settings affect performance; suboptimal configurations can impact results.
• In rare cases, outputs may contain hallucinated or contextually misaligned information.

Training Details

Parameter	Value
Dataset Size	108K samples (Modular Combustion of Datasets)
Model Architecture	Qwen2_5_VLForConditionalGeneration
Total Disk Volume	300,000 MB
Training Time	approx. 12,897 seconds (~3.58 hours)
Warmup Steps	750
Precision	bfloat16

References

• DocVLM: Make Your VLM an Efficient Reader https://arxiv.org/pdf/2412.08746v1

• YaRN: Efficient Context Window Extension of Large Language Models
  https://arxiv.org/pdf/2309.00071

• Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution
  https://arxiv.org/pdf/2409.12191

• Qwen-VL: A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyond
  https://arxiv.org/pdf/2308.12966

• A Comprehensive and Challenging OCR Benchmark for Evaluating Large Multimodal Models in Literacy https://arxiv.org/pdf/2412.02210