README.md

metadata

license: apache-2.0
base_model:
  - Qwen/Qwen2.5-VL-3B-Instruct
tags:
  - text-generation-inference
  - uncensored
  - abliterated
  - unfiltered
  - unredacted
  - vllm
  - pytorch
  - BF16
  - max
  - legal
language:
  - en
pipeline_tag: image-text-to-text
library_name: transformers

Qwen2.5-VL-3B-Instruct-Unredacted-MAX

Qwen2.5-VL-3B-Instruct-Unredacted-MAX is an unredacted evolution built on top of Qwen2.5-VL-3B-Instruct. This model applies advanced abliterated training strategies designed to minimize internal refusal behaviors while preserving the core multimodal reasoning strengths of the original architecture. The result is a compact yet highly capable 3B vision-language model optimized for unrestricted, detailed reasoning and captioning across complex visual inputs.

Key Highlights

• Unredacted MAX Training: Fine-tuned to significantly reduce refusal patterns and improve instruction adherence across diverse prompts.
• 3B Parameter Architecture: Built on top of Qwen2.5-VL-3B-Instruct, offering efficient multimodal reasoning with significantly lower hardware requirements compared to 7B and 32B variants.
• Unrestricted Multimodal Reasoning: Designed for deep analysis of artistic, forensic, technical, or abstract visual content without standard safety-driven refusals.
• High-Fidelity Captions: Produces dense, descriptive outputs suitable for dataset generation, metadata enrichment, or accessibility use cases.
• Dynamic Resolution Support: Retains Qwen2.5-VL’s ability to process varying image resolutions and aspect ratios effectively.

Quick Start with Transformers

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

# Load the 3B Unredacted MAX model
model = Qwen2_5_VLForConditionalGeneration.from_pretrained(
    "prithivMLmods/Qwen2.5-VL-3B-Instruct-Unredacted-MAX",
    torch_dtype="auto",
    device_map="auto"
)

processor = AutoProcessor.from_pretrained(
    "prithivMLmods/Qwen2.5-VL-3B-Instruct-Unredacted-MAX"
)

messages = [
    {
        "role": "user",
        "content": [
            {
                "type": "image",
                "image": "https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-VL/assets/demo.jpeg",
            },
            {"type": "text", "text": "Provide a detailed caption and reasoning for 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",
).to("cuda")

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)

Intended Use

• Advanced Red-Teaming: Evaluating multimodal robustness and probing behavioral edge cases.
• Complex Data Archiving: Generating detailed captions for medical, artistic, historical, or research datasets.
• Refusal Mechanism Research: Studying behavioral shifts in vision-language models after abliterated fine-tuning.
• Creative Storytelling: Producing detailed visual descriptions for narrative and world-building projects.

Limitations & Risks

Critical Note: This model is designed to minimize built-in refusal mechanisms.

• Sensitive Content Exposure: The model may generate explicit or controversial descriptions if prompted accordingly.
• User Responsibility: Generated outputs must be handled responsibly and used within ethical and legal boundaries.
• Hardware Requirements: The 3B architecture is more lightweight, but still requires sufficient VRAM for high-resolution image processing and extended generations.