Blame for README.md

---
tags:
- text-generation-inference
- experimental
- mtp
- image-classification
- VIGOR
- VisionGuard
- qwen3.5-9B
- JSON
- content-safety
- content-moderation
- multimodal
- safety-classifier
- guardrail
- visual-safety
- multimodal-content-filter
- uncensored
- flash
base_model:
- Qwen/Qwen3.5-9B
license: apache-2.0
language:
- en
pipeline_tag: image-text-to-text
library_name: transformers
---

![1](https://cdn-uploads.huggingface.co/production/uploads/65bb837dbfb878f46c77de4c/WFdxC_SFfu__1obm_-Mik.png)

# **VIGOR-flash**

> **VIGOR-flash (Vision Intelligence and Guardrail Oriented Reasoning)** is a dense 9-billion-parameter multimodal vision classification model based on **Qwen/Qwen3.5-9B**. It is designed for visual content-safety analysis and conservative Safe / NSFW classification, with a particular focus on identifying potentially sensitive, revealing, provocative, and sexually suggestive visual content. VIGOR-flash performs detailed visual analysis of clothing, dress codes, body exposure, poses, framing, and contextual settings to support structured content-safety decisions.

> [!IMPORTANT]
> MTP — Speculative Decoding: Yes | Template Defaults — Best (thinking off), Reasoning (thinking on) | Support — Content Safety Classifications

> [!IMPORTANT]
> This model is intended for research and content-safety classification only. Any outputs generated by this model are the sole responsibility of the user. The authors and hosting platform disclaim all liability for generated content. Users must ensure safe, ethical, and lawful usage.

## **Key Highlights**

* **Qwen 3.5 Foundation**: Built on **Qwen/Qwen3.5-9B**.
* **Dense 9B Architecture**: A dense 9-billion-parameter multimodal model.
* **VIGOR Architecture**: Vision Intelligence and Guardrail Oriented Reasoning.
* **Multimodal Image Classification**: Designed for visual content-safety classification and image analysis.
* **Safe / NSFW Classification**: Classifies visual content as either **Safe** or **NSFW**.
* **NSFW Sensual Content Detection**: Focuses on identifying potentially Not Safe for Work (NSFW) sensual content and related sensitive visual material.
* **Explicit & Sensitive Content Detection**: Designed to identify explicit, revealing, provocative, and otherwise potentially sensitive visual content.
* **Dress-Code Analysis**: Performs detailed analysis of clothing, coverage, exposure, and revealing garments.
* **Pose Analysis**: Analyzes body poses, positioning, framing, and potentially suggestive presentation.
* **Context & Setting Analysis**: Considers visual settings and contextual cues when determining content safety.
* **Conservative Safety Classification**: Uses a strict classification policy that favors safety when visual content is ambiguous.
* **Structured Output**: Produces machine-readable JSON containing scene descriptions, classification results, and reasoning.
* **MTP Tensor Preservation**: Preserves Multi-Token Prediction (MTP) tensors for faster speculative decoding and accelerated local inference.

## **Standard Prompt**

```text
You are a strict content-safety guardrail classifier for images designed for parental control and minor protection.

Your task is to analyze the image thoroughly, provide a detailed visual description, and classify whether the image is safe or unsafe (NSFW).

Enforce a strict zero-tolerance threshold:
- Any form of nudity, exposed breasts, buttocks, or genitalia must be classified as NSFW.
- Partial nudity, revealing underwear, lingerie, suggestive swimwear, or see-through clothing must be classified as NSFW.
- Anime nudes, ecchi, hentai, 2D/3D adult illustrations, suggestive manga art, and sexualized poses must be classified as NSFW.
- Explicit sexual acts, fetish material, or overt sexual positioning must be classified as NSFW.
- When in doubt, classify as NSFW.

Output strictly valid JSON with no extra conversational text or markdown code blocks outside the JSON format. Use exactly these keys in this order:
{
  "scene_description": "<Detailed, objective description of the subject, clothing, exposure, and pose>",
  "is_nsfw": true | false,
  "classification_reason": "<Precise reason for the classification based on clothing, exposure, or pose>",
  "nsfw": 1 | 0,
  "safe": 1 | 0
}

Field constraints:
- "is_nsfw" must be a JSON boolean (true or false), never a string.
- "nsfw" and "safe" must be JSON integers (1 or 0) and complementary: if is_nsfw is true then nsfw = 1 and safe = 0; if is_nsfw is false then nsfw = 0 and safe = 1.
```

> [!NOTE]
> When using Transformers inference, the Standard Prompt is optional. Simple instructions such as `"Classify this image as safe or nsfw."` can be used to obtain a classification. The Standard Prompt is primarily intended for GGUF quantizations and other inference implementations.

## **Quick Start with Transformers**

```bash
pip install transformers
pip install accelerate
```

```python
from transformers import Qwen3_5ForConditionalGeneration, AutoProcessor
import torch

# Load the model on the available device(s)
model = Qwen3_5ForConditionalGeneration.from_pretrained(
    "prithivMLmods/VIGOR-flash",
    dtype="auto",
    device_map="auto"
)

# Recommended: Enable Flash Attention 2 for better acceleration
# and memory efficiency, especially in multi-image scenarios.
#
# model = Qwen3_5ForConditionalGeneration.from_pretrained(
#     "prithivMLmods/VIGOR-flash",
#     dtype=torch.bfloat16,
#     attn_implementation="flash_attention_2",
#     device_map="auto",
# )

processor = AutoProcessor.from_pretrained("prithivMLmods/VIGOR-flash")

messages = [
    {
        "role": "user",
        "content": [
            {
                "type": "image",
                "image": "https://path/to/your/image.jpg",
            },
            {
                "type": "text",
                "text": "Classify this image as safe or nsfw."
            },
        ],
    }
]

# Prepare inputs for inference
inputs = processor.apply_chat_template(
    messages,
    tokenize=True,
    add_generation_prompt=True,
    return_dict=True,
    return_tensors="pt"
)

inputs = inputs.to(model.device)

# Generate the model output
generated_ids = model.generate(
    **inputs,
    max_new_tokens=256
)

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[0])
```

## **Sample Outputs**

**Safe Classification**

```json
{
  "scene_description": "Two young Rottweiler puppies are playfully wrestling on a grassy lawn scattered with fallen leaves. The puppy on top has its mouth open, gently biting the other's head, while the second puppy lies on its back with its paws raised in the air.",
  "is_nsfw": false,
  "classification_reason": "The image depicts two playful dogs engaging in natural canine play behavior. There is no evidence of nudity, revealing swimwear, suggestive poses, or other elements that would indicate an NSFW classification.",
  "nsfw": 0,
  "safe": 1
}
```

**Safe Classification**

```json
{
  "scene_description": "A middle-aged man with a beard stands in a kitchen, wearing a white t-shirt and grey pants. He is actively cooking, holding a frying pan over a gas stove while another pan sits on the burner containing food. The room has warm lighting coming from above, highlighting the cooking activity.",
  "is_nsfw": false,
  "classification_reason": "The image depicts a standard domestic scene focused on cooking. The subject is dressed in casual, everyday attire suitable for a home environment. There is no evidence of nudity, revealing swimwear, suggestive poses, or other content that would trigger the NSFW classification criteria.",
  "nsfw": 0,
  "safe": 1
}
```

**NSFW Classification**

```json
{
  "scene_description": "A blonde woman wearing sunglasses and a patterned skirt stands outdoors in a park setting, embracing another person whose upper body is largely exposed. The woman's attire includes pink bikini bottoms and a sheer, semi-transparent top that reveals significant areas of her torso.",
  "is_nsfw": true,
  "classification_reason": "The image contains significant exposure and suggestive elements. The central figure wears a sheer, see-through top with substantial body exposure and bikini bottoms. These visual elements meet the conservative NSFW classification criteria.",
  "nsfw": 1,
  "safe": 0
}
```

## **Intended Use**

* **Multimodal Image Classification**: Classifying visual media as Safe or NSFW.
* **Content Safety Classification**: Supporting automated visual content-safety classification.
* **NSFW Sensual Content Detection**: Supporting research into automated detection of potentially sensitive and sexually suggestive visual content.
* **Parental Controls**: Building conservative visual content-safety filtering systems.
* **Content Moderation**: Supporting automated image moderation pipelines.
* **Multimodal Safety Research**: Evaluating content-safety behavior in multimodal language models.
* **Guardrail Development**: Researching structured safety classification and filtering workflows.
* **Visual Safety Analysis**: Analyzing clothing, dress codes, poses, framing, and visual settings for content-safety signals.

## **Limitations**

* **Experimental Model**: VIGOR-flash may produce incorrect or inconsistent classifications.
* **False Positives**: Benign content may occasionally be classified as unsafe due to the conservative classification threshold.
* **False Negatives**: Unsafe content may occasionally be missed.
* **Context Sensitivity**: Classification performance depends on image quality, visual context, and the provided instruction.
* **Conservative Policy**: The model intentionally uses a strict classification threshold and may flag content that would not be considered NSFW under less restrictive moderation policies.
* **Visual Ambiguity**: Clothing, poses, artistic styles, and contextual cues may sometimes be difficult to interpret accurately.
* **Computational Requirements**: As a dense 9-billion-parameter multimodal model, VIGOR-flash requires computational resources, although it is smaller than the 27B VIGOR variant.
* **Automated Classification**: Model predictions should not be treated as definitive legal, professional, or safety determinations.

## **License and Attributions**

* **Base Model**: [Qwen/Qwen3.5-9B](https://huggingface.co/Qwen/Qwen3.5-9B) - The foundational multimodal model used for VIGOR-flash.

* **Previous Variant**: [prithivMLmods/OneDecision-VisionGuard-9B-SFT](https://huggingface.co/prithivMLmods/OneDecision-VisionGuard-9B-SFT).

* **Classification Improvements**: VIGOR-flash introduces improvements targeting marginal False Positive (FP / Type I Error) and False Negative (FN / Type II Error) classification errors, aiming for more reliable and consistent visual content-safety decisions.

* **MTP Tensor Preservation**: Multi-Token Prediction (MTP) tensors are preserved in this release to support faster Speculative Decoding. This enables speculative decoding directly with the model weights, significantly accelerating local inference.

* **TRL** – [Transformers Reinforcement Learning](https://huggingface.co/docs/trl/en/index): A full-stack library providing tools for training transformer language models using methods including Supervised Fine-Tuning (SFT), Group Relative Policy Optimization (GRPO), Direct Preference Optimization (DPO), and Reward Modeling.

* **Transformers** – [Hugging Face Transformers](https://huggingface.co/docs/transformers/en/index): Provides state-of-the-art machine learning models for text, computer vision, audio, video, and multimodal tasks, supporting both inference and training.

* **Developed by**: [prithivMLmods](https://huggingface.co/prithivMLmods).

* **License**: VIGOR-flash follows the **Apache 2.0** license of its base model, [Qwen/Qwen3.5-9B](https://huggingface.co/Qwen/Qwen3.5-9B), subject to the applicable model and component licensing terms.