Brello EI 0 - Emotional Intelligence AI Model

Created by Epic Systems | Engineered by Rehan Temkar

A locally-run emotional intelligence AI model designed to provide empathetic, emotionally-aware responses with natural conversation flow.

Features

Emotional Intelligence: Designed to provide empathetic, understanding responses
Local Operation: Runs completely locally without external dependencies
Memory Efficient: 4-bit quantization for optimal performance on limited hardware
Advanced Architecture: Based on Llama 3.2 3B foundation model
Easy Integration: Simple API for quick integration
Flexible Configuration: Customizable generation parameters

Installation

Prerequisites

Python 3.8+
CUDA-compatible GPU (recommended) or CPU
At least 8GB RAM (16GB recommended)

Install Dependencies

pip install -r requirements.txt

Model Options

Option 1: Use Public Model (Recommended for quick start) The default configuration uses microsoft/DialoGPT-medium which is publicly available and doesn't require authentication.

Option 2: Use Llama 3.2 3B (Requires authentication) To use the actual Llama 3.2 3B model:

Create a Hugging Face account
Accept the model license at: https://huggingface.co/meta-llama/Llama-3.2-3B-Instruct
Login with: huggingface-cli login or hf auth login
Update the model_path in your code to: "meta-llama/Llama-3.2-3B-Instruct"

Option 3: Use Other Public Models

microsoft/DialoGPT-large (larger, better responses)
microsoft/DialoGPT-small (faster, smaller)
HuggingFaceTB/SmolLM3-3B (3B parameter model)

Quick Start

Basic Usage

from brello_ei_0 import BrelloEI0

# Load the model
model = BrelloEI0(
    model_path="microsoft/DialoGPT-medium",  # Public model, no auth required
    load_in_4bit=False  # Set to True if you have CUDA
)

# Generate an emotionally intelligent response
response = model.generate_response("I'm feeling really stressed about my job interview.")
print(response)

Alternative Loading

from brello_ei_0 import load_brello_ei_0

# Load model using convenience function
model = load_brello_ei_0("microsoft/DialoGPT-medium")

# Direct call
response = model("I'm really happy about my recent success!")
print(response)

Chat Interface

# Simple chat
response = model.chat("How are you feeling today?")
print(response)

Example Conversations

# Example 1: Anxiety Support
response = model.generate_response("I'm feeling really anxious about my presentation tomorrow.")
# Output: "I can understand how nerve-wracking presentations can be. It's completely natural to feel anxious..."

# Example 2: Celebrating Success
response = model.generate_response("I just got promoted at work!")
# Output: "That's wonderful! I can feel your excitement and it's absolutely contagious..."

# Example 3: Emotional Support
response = model.generate_response("I'm feeling lonely and isolated.")
# Output: "I'm so sorry you're feeling this way. Loneliness can be really painful..."

# Example 4: Career Guidance
response = model.generate_response("I'm confused about what I want to do with my life.")
# Output: "That's a really common and natural feeling, especially when we're at crossroads..."

Configuration

Model Parameters

model_path: Path to Llama 3.2 3B model (default: "meta-llama/Meta-Llama-3.2-3B-Instruct")
device: Device to load model on ('cuda', 'cpu', etc.)
load_in_4bit: Enable 4-bit quantization for memory efficiency (recommended)
load_in_8bit: Enable 8-bit quantization for memory efficiency
torch_dtype: Torch data type for model weights

Generation Parameters

temperature: Sampling temperature (default: 0.7)
top_p: Top-p sampling parameter (default: 0.9)
max_length: Maximum response length (default: 4096)
min_length: Minimum response length (default: 30)
max_new_tokens: Maximum new tokens to generate (default: 256)
repetition_penalty: Penalty for repetition (default: 1.1)

Performance

Model Specifications

Foundation: Microsoft DialoGPT-medium
Parameters: 345 Million
Context Length: 1024 tokens
Training: Conversational dialogue data
Optimization: Emotional intelligence focus

Memory Requirements

Full Precision: ~1GB VRAM
8-bit Quantization: ~500MB VRAM
4-bit Quantization: ~250MB VRAM (recommended)

Advanced Usage

Custom Generation Parameters

response = model.generate_response(
    "I'm feeling overwhelmed with my responsibilities.",
    temperature=0.8,
    top_p=0.95,
    max_new_tokens=300,
    repetition_penalty=1.05
)

Batch Processing

messages = [
    "I'm really proud of my accomplishments.",
    "I'm feeling uncertain about my future.",
    "I'm grateful for my support system."
]

responses = []
for message in messages:
    response = model.generate_response(message)
    responses.append(response)

🎯 Training

Fine-tune for Emotional Intelligence

python train_brello_ei_0.py

The training script will:

Load Llama 3.2 3B with 4-bit quantization
Apply LoRA for efficient fine-tuning
Train on emotional intelligence data
Save the fine-tuned model

Training Data

The model is fine-tuned on emotional intelligence scenarios:

Anxiety and stress support
Celebrating success and achievements
Dealing with loneliness and isolation
Career guidance and life decisions
Gratitude and appreciation
Overwhelm and responsibility management

Architecture

Brello EI 0 is built on advanced language model architecture with the following key components:

Base Model: Microsoft DialoGPT-medium
Tokenizer: Optimized for conversational data
Generation: Emotionally intelligent response patterns
Post-processing: Response cleaning and enhancement
Quantization: 4-bit for memory efficiency (optional)

🎯 Use Cases

Emotional Support

Providing empathetic responses to stress and anxiety
Supporting users through difficult life transitions
Celebrating achievements and successes

Personal Development

Career guidance and decision-making support
Life goal exploration and planning
Self-reflection and emotional awareness

Mental Health Support

Stress management and coping strategies
Emotional validation and understanding
Positive reinforcement and encouragement

Contributing

This model is part of the Epic Systems AI initiative. For questions or contributions, please contact the development team.

License

This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgments

Epic Systems for the vision and support
Rehan Temkar for engineering and development
Microsoft for the DialoGPT foundation model
Hugging Face for the transformers library

Brello EI 0 - Bringing emotional intelligence to AI conversations 💙✨