---
license: apache-2.0
language:
- en
base_model:
- EssentialAI/rnj-1-instruct
base_model_relation: finetune
library_name: peft
tags:
- tactical-intelligence
- vanta-research
- scout-8b
- conversational
- conversational-ai
- research
- roleplay
- chat
- chat-ai
- chatbot
- reasoning
- logic
- science
- STEM
- constraint-aware
---

<div align="center">

![vanta_trimmed](https://cdn-uploads.huggingface.co/production/uploads/686c460ba3fc457ad14ab6f8/hcGtMtCIizEZG_OuCvfac.png)
  
  <h1>VANTA Research</h1>
    
  <p><strong>Independent AI research lab building safe, resilient language models optimized for human-AI collaboration</strong></p>
  
  <p>
    <a href="https://vantaresearch.xyz"><img src="https://img.shields.io/badge/Website-vantaresearch.xyz-black" alt="Website"/></a>
    <a href="https://merch.vantaresearch.xyz"><img src="https://img.shields.io/badge/Merch-merch.vantaresearch.xyz-sage" alt="Merch"/></a>
    <a href="https://x.com/vanta_research"><img src="https://img.shields.io/badge/@vanta_research-1DA1F2?logo=x" alt="X"/></a>
    <a href="https://github.com/vanta-research"><img src="https://img.shields.io/badge/GitHub-vanta--research-181717?logo=github" alt="GitHub"/></a>
  </p>
</div>

---



# Scout-8B: Tactical Intelligence for Edge Devices

**VANTA Research Entity-002: The Reconnaissance Specialist**

## Overview

Scout-8B is a specialized AI model built for tactical intelligence and reconnaissance operations. Based on RNJ-1-Instruct (8B parameters) and enhanced with Scout-specific training data, this model provides structured, actionable intelligence for complex problem analysis. This model contains all of the same data used in [Scout-4B](https://huggingface.co/vanta-research/scout-4b) - and not only improves, but expands on previous capabilities. 


## Capabilities

**Tactical Intelligence Analysis**
- Systematic problem decomposition
- Structured reconnaissance approach
- Data-driven assessment methodology

**Operational Planning** 
- Multi-phase operation planning
- Risk assessment and mitigation
- Resource allocation guidance

**Technical Assessment**
- Architecture evaluation and analysis
- Performance optimization recommendations
- Security perimeter assessment

## Usage

Scout is design and optimized for the following styles of interaction: 
- Direct, professional communication style
- Tactical terminology usage
- Structured, phased approaches
- Focus on actionable outcomes

## Files

- `config.json` - Model configuration
- `scout_config.json` - Scout-specific settings
- `model-*.safetensors` - Merged model weights
- `tokenizer.*` - Tokenizer files

## Training Details

- **Method**: LoRA (Low-Rank Adaptation)
- **Rank**: 16, Alpha: 32
- **Epochs**: 2
- **Dataset**: ~4,500 Scout-specific synthetically generated examples
- **Target Modules**: q_proj, k_proj, v_proj, o_proj, gate_proj, up_proj, down_proj

## Next Steps

1. **Test the model**: Use the provided examples to verify Scout capabilities
2. **Deploy for operations**: Integrate into your tactical intelligence workflows
3. **Customize further**: Additional fine-tuning for specific operational contexts

## Contact
- Organization: hello@vantaresearch.xyz
- Engineering/Design: tyler@vantaresearch.xyz

*Proudly developed by VANTA Research in Portland, Oregon*