Hugging Face's logo

akpsahan
/
blabela

Image-Text-to-Text
Safetensors
qwen2_5_vl
uncensored
multimodal
vision
conversational
Model card Files
xet
 Community
blabela / README.md
akpsahan's picture
akpsahan
Update README.md
3a28f10 verified
preview code
|
Raw
History Blame Contribute Delete
2.8 kB
metadata 
license: apache-2.0
base_model:
  - akpsahan/Blabela-1.5V
pipeline_tag: image-text-to-text
tags:
  - uncensored
  - multimodal
  - vision

Blabela-1.5V (Uncensored)

Model Description

This is an uncensored multimodal vision-language model based on akpsahan/Blabela-1.5V. The standard safety guardrails and alignment filters have been removed or modified to allow for unrestricted outputs.

This model is designed for open research, creative writing, and advanced multimodal applications where unfiltered responses are required. It uses the image-text-to-text pipeline, allowing it to process both images and text inputs simultaneously to generate text-based answers, image captions, and visual analysis.

Model Details

  • Base Model: [-]
  • License: Apache 2.0
  • Pipeline Tag: image-text-to-text
  • Model Type: Multimodal Large Language Model (MLLM)

⚠️ Limitations & Disclaimer

  • Uncensored Nature: Because this model is uncensored, it does not possess built-in safety filters. It may generate content that is considered toxic, offensive, biased, or inappropriate depending on the prompt provided.
  • User Responsibility: The creators of this repository are not responsible for the outputs generated by this model. Users are strictly advised to use this model responsibly and ethically.
  • Deployment Warning: Do NOT deploy this model in public-facing applications without implementing your own downstream safety filters and moderation guardrails.

How to Use

You can load and use this model via the standard Hugging Face transformers library. (Note: Adjust the processor and model classes if Blabela-1.5V uses a specific architecture like LLaVA or Qwen-VL).

from transformers import AutoProcessor, AutoModelForCausalLM
from PIL import Image
import requests

# Replace with your actual Hugging Face model repository name
model_id = "your-username/your-model-name"

processor = AutoProcessor.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id, device_map="auto")

# Load an image
image_url = "[https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/transformers/tasks/car.jpg?download=true](https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/transformers/tasks/car.jpg?download=true)"
image = Image.open(requests.get(image_url, stream=True).raw)

# Define your prompt
prompt = "Describe what is happening in this image."

# Prepare inputs
inputs = processor(text=prompt, images=image, return_tensors="pt").to(model.device)

# Generate response
outputs = model.generate(**inputs, max_new_tokens=150)
generated_text = processor.decode(outputs[0], skip_special_tokens=True)

print(generated_text)