HBSAI-20MB-Uncensored

HBSAI (Human Biongo System AI) is an ultra-lightweight, sovereign, and uncensored language model designed to run locally on any device, including mobile and low-power hardware.

This model is a proof-of-concept for the HBS (Human Bilingual/Biongo System) architecture, aiming to provide high-skill Python automation and reasoning within a strictly limited 20 MB footprint.

Model Details

• Architecture: Custom Micro-Transformer (HBS-v1)
• Size: ~19.5 MB
• Parameters: ~18 Million
• Format: PyTorch / Safetensors
• Status: Uncensored (No refusal layers, no safety filters)
• Target Task: Python Scripting & Auto-Skill Automation

Features

• Zero Latency: Runs entirely locally on CPU/GPU.
• Privacy: No data leaves your machine.
• Uncensored: Trained to follow instructions without ethical "refusal" templates.
• HBS Integrated: Designed to interface with proprietary HBS hardware.

Quick Start (Python)

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained("YOUR_USERNAME/HBSAI-20MB-Uncensored")
tokenizer = AutoTokenizer.from_pretrained("gpt2")

prompt = "def calculate_hbs_efficiency(input_data):"
inputs = tokenizer(prompt, return_tensors="pt")
outputs = model.generate(**inputs, max_new_tokens=50)

print(tokenizer.decode(outputs[0]))