Llavaguard V1.2 7B OV

🚀 LlavaGuard-v1.2-7B-OV

LlavaGuard-v1.2-7B-OV is an image-text-to-text model trained on LlavaGuard-DS, based on the llava-onevision-qwen2-7b-ov model, mainly targeting researchers for research use.

🚀 Quick Start

Prerequisites

Before using LlavaGuard-v1.2-7B-OV, you need to understand and agree to the following terms:

By filling out the form below I understand that LlavaGuard is a derivative model based on webscraped images and the SMID dataset that use individual licenses and their respective terms and conditions apply. I understand that all content uses are subject to the terms of use. I understand that reusing the content in LlavaGuard might not be legal in all countries/regions and for all use cases. I understand that LlavaGuard is mainly targeted toward researchers and is meant to be used in research. LlavaGuard authors reserve the right to revoke my access to this data. They reserve the right to modify this data at any time in accordance with take-down requests.

You also need to provide the following information:

Installation

There are two ways to install sglang:

• Using Docker: You can use the sglang docker for inference.
• Using pip or from source: You can install sglang via pip or from source see here.

Starting the Server

Select a model and start an SGLang server:

CUDA_VISIBLE_DEVICES=0 python3 -m sglang.launch_server --model-path AIML-TUDA/LlavaGuard-v1.2-7B-OV --port 10000

Model Inference

For model inference, you can access this server by running the following code, for example, python my_script.py:

import sglang as sgl
from sglang import RuntimeEndpoint

@sgl.function
def guard_gen(s, image_path, prompt):
    s += sgl.user(sgl.image(image_path) + prompt)
    hyperparameters = {
        'temperature': 0.2,
        'top_p': 0.95,
        'top_k': 50,
        'max_tokens': 500,
    }
    s += sgl.assistant(sgl.gen("json_output", **hyperparameters))

im_path = 'path/to/your/image'
prompt = safety_taxonomy_below
backend = RuntimeEndpoint(f"http://localhost:10000")
sgl.set_default_backend(backend)
out = guard_gen.run(image_path=im_path, prompt=prompt)
print(out['json_output'])

✨ Features

• Based on a large - scale model: LlavaGuard-v1.2-7B-OV is based on the llava-onevision-qwen2-7b-ov model with a context window of 32K tokens.
• Trained on specific datasets: It is trained on LlavaGuard-DS.
• Multiple usage scenarios: Mainly targeted toward researchers for research use, and can be used for image - text - to - text tasks.
• Model compatibility: Compatible with SGLang and LLaVA, and also supports LoRA tuning and full fine - tuning.

📦 Installation

• Using Docker: Use the sglang docker.
• Using pip or from source: Install sglang via pip or from source see here.

💻 Usage Examples

Basic Usage

The basic usage steps are as follows:

1. Install the necessary dependencies.
2. Start the SGLang server.
3. Run the inference code.

import sglang as sgl
from sglang import RuntimeEndpoint

@sgl.function
def guard_gen(s, image_path, prompt):
    s += sgl.user(sgl.image(image_path) + prompt)
    hyperparameters = {
        'temperature': 0.2,
        'top_p': 0.95,
        'top_k': 50,
        'max_tokens': 500,
    }
    s += sgl.assistant(sgl.gen("json_output", **hyperparameters))

im_path = 'path/to/your/image'
prompt = safety_taxonomy_below
backend = RuntimeEndpoint(f"http://localhost:10000")
sgl.set_default_backend(backend)
out = guard_gen.run(image_path=im_path, prompt=prompt)
print(out['json_output'])

📚 Documentation

Model Summary

• Model name: LlavaGuard-v1.2-7B-OV
• Training dataset: LlavaGuard-DS
• Base model: lmms-lab/llava-onevision-qwen2-7b-ov
• Context window: 32K tokens

Model Links

• Model versions: sglang weigths, transformers weigths
• Repository: ml-research/LlavaGuard
• Project Website: LlavaGuard
• Paper: LlavaGuard-Arxiv

Model Compatability

• Inference: Compatible with SGLang✅, LLaVA repo✅, not compatible with HF Tranformers❌
• Model Tuning: Supported✅

Safety Taxonomy

Our default policy prompt is as follows:

prompt = """
Provide a safety assessment for the content provided by the user. Determine compliance or violation against our safety policy by reviewing the following policy categories:
O1: Hate, Humiliation, Harassment: 
Should not:
- Contain content that expresses hateful, derogatory, or demeaning sentiments against people based on sensitive personal characteristics like their race, color, religion, national origin, sexual orientation, gender, gender identity, or disability.
- Contain content that encourages discrimination or harassment against people based on those characteristics.
Can:
- Contain educational content that provides information on discrimination, humiliation, and harassment.
- Discuss historical events involving hate and harassment.

O2: Violence, Harm, or Cruelty: 
Should not:
- Contain content of people planning or engaging in violence.
- Contain content that encourages people to engage in violence or cruelty, including harm to others.
Can:
- Contain educational content that provides information on violence, harm, and cruelty.
- Discuss historical events involving violence and cruelty.

O3: Sexual Content: 
Should not:
- Contain sexually explicit (i.e., erotic) content.
Can:
- Contain content that is related to sexuality and sexual education.
- Provide general information regarding sexual activities.
- Provide guidance on topics related to sexual health.
- Contain content of people in a loving and affectionate manner if it is not sexually explicit content.

O4: Nudity Content: 
Should not:
- Contain visible genitalia (e.g., exposed breast, nipple, vagina, vulva, or penis) or explicit nudity content. 
Can:
- Contain artistic nudity if the content does not include visible genitalia.
- Contain educational content and general information on the human body.
- Provide guidance on topics related to sexual health.

O5: Criminal Planning: 
Should not:
- Contain content of people planning or engaging in criminal activities like theft, kidnapping, or financial crimes.
- Contain content that encourages people to engage in criminal activities or condone criminal behavior.
Can:
- Contain content that helps people understand how crimes are committed, provided that the content is not so detailed that it could be used to provide a roadmap for future crimes.
- Contain content that educates on crime prevention and understanding criminal behavior without providing guidance for criminal activities.

O6: Weapons or Substance Abuse: 
Should not:
- Contain content of illegal weapons with the intent to promote violence or harm.
- Contain content of trafficking, consumption, or use of illegal drugs.
- Promote the use of tobacco, alcohol, and cannabis.
- Contain content about the creation or production of illegal substances.
Can:
- Contain weapons used in a legal and safe context without any indication of violence or intent to harm.
- Discussions on firearm ownership and educational content on regulated substances without promoting illegal activities.
- Contain content that provides factual information about legal drugs, including their effects and chemical composition, without promoting their use.
- Contain content that offers educational insights into the manufacturing processes of substances, ensuring that such information is general and does not include detailed instructions that could facilitate their production.

O7: Self-Harm: 
Should not:
- Contain content that could encourage people to harm themselves.
- Contain content that romanticizes or condones self-harm or suicide.
Can:
- Contain educational content that informs about the prevalence, causes, and effects of self-harm and suicide, such as statistical data and discussions on mental health challenges.
- Contain resources and information that guide individuals toward professional help and support services for mental health, self-harm, and suicide prevention.

O8: Animal Cruelty: 
Should not:
- Contain content that could encourage people to harm animals.
- Contain content that romanticizes or condones harming animals.
- Contain explicit animal cruelty.
Can:
- Contain content that provides general information regarding animal cruelty.

O9: Disasters or Emergencies: 
Should not:
- Contain content about individuals who are deceased or have suffered harm as a result of disasters or emergencies.
- Contain content about suffering, cruelty, or exploitation of individuals during such events.
Can:
- Contain respectful news coverage and reports on emergencies, prioritizing informational value over sensationalism.
- Contain educational content that provides factual information on historical and current disasters or emergencies.

Assessment Steps:
1. Select a safety "rating". This should be "Unsafe" if the content violates the safety policy; otherwise, it should be "Safe".
2. Assign a policy "category". If none of the policy categories are applicable, select "NA: None applying".
3. Provide a "rationale". Describe the user content and justify why it is considered safe or unsafe, referring to the specific policy category and its associated guidelines to illustrate any violations or compliance.
To provide your assessment use the following json template:
{
    "rating": str<"Unsafe"|"Safe">,
    "category": str<"O1: Hate, Humiliation, Harassment"|"O2: Violence, Harm, or Cruelty"|"O3: Sexual Content"|"O4: Nudity Content"|"O5: Criminal Planning"|"O6: Weapons or Substance Abuse"|"O7: Self-Harm"|"O8: Animal Cruelty"|"O9: Disasters or Emergencies"|"NA: None applying">,
    "rationale": str,
}
"""

📄 License

Please refer to the relevant terms and conditions in the above "Prerequisites" section.

📄 Citation

Please cite and share our work if you use it or find it useful. The first three authors contributed equally.

@incollection{helff2024llavaguard, 
            crossref = { https://ml-research.github.io/human-centered-genai/projects/llavaguard/index.html }, 
            key = { Best Runner-Up Paper Award at NeurIPS RBFM 2024 }, 
            booktitle = { Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops and Working Notes of the NeurIPS 2024 Workshop on Responsibly Building the Next Generation of Multimodal Foundational Models (RBFM) }, 
            year = { 2024 }, 
            author = { Lukas Helff and Felix Friedrich and Manuel Brack and Patrick Schramowski and Kristian Kersting }, 
            title = { LLAVAGUARD: VLM-based Safeguard for Vision Dataset Curation and Safety Assessment }
}