GitHub - MongooseReborn/kalos-engine: Project Kalos: A Biological-Resonant C/CUDA Neuromorphic Engine Suite & Empirical Whitepapers v1.0 · GitHub
/" data-turbo-transient="true" />
Skip to content
Type / to search
Sign in<br>Sign upAppearance settings
You signed in with another tab or window. Reload to refresh your session.<br>You signed out in another tab or window. Reload to refresh your session.<br>You switched accounts on another tab or window. Reload to refresh your session.
Dismiss alert
{{ message }}
MongooseReborn
kalos-engine
Public
Notifications<br>You must be signed in to change notification settings
Fork
Star
main
BranchesTags
Go to file
CodeOpen more actions menu
Folders and files<br>NameNameLast commit message<br>Last commit date<br>Latest commit
History<br>1 Commit<br>1 Commit
bin
bin
docs
docs
Modelfile
Modelfile
README.md
README.md
kalos_monitor.py
kalos_monitor.py
kalos_sms.py
kalos_sms.py
View all files
Repository files navigation
license<br>mit
language
en
tags
cuda
neuromorphic
spiking-neural-network
associative-memory
llm-acceleration
vram-optimization
c-cpp
pipeline_tag<br>text-generation
library_name<br>c-cuda
extra_gated_heading<br>Project Kalos C/CUDA Neuromorphic Suite
🌿 Project Kalos: High-Performance C/CUDA Neuromorphic Engine Suite
Project Kalos is a high-performance, biological-resonant C/CUDA neuromorphic engine suite designed to decouple long-range dialogue memory, spiking neural reflexes, and physical sensory haptics from large language model (LLM) text tokenization.
💥 Key Features & Performance Highlights
O(1) Microsecond Memory Recall (465.12 μs): Replaces linear text re-tokenization (20,441.90 ms) with constant-time CUDA vector lookup.
7.25x Faster Total Response Completion (72B Models): Cuts total user-sent to output-completion latency on 72B parameter models from 23.71 seconds down to 3.27 seconds .
99.8% VRAM Footprint Reduction: Compresses 16.38 GB KV-cache bloat down to 2.50 MB of sparse associative neural templates.
Sub-Millisecond Reflex Sentry (15.16 μs): 4.19-Million CUDA spiking neurons for instant event detection.
Native Physical Haptics: Real-time CUDA perception for physical touch contact, localized warmth, and FFT audio spectrum resonance.
Stateful Identity Persistence (.soul): Compact 2.5 MB binary cortical persistence format.
🛠️ Quick Start & Running Precompiled Release Binaries
1. Clone the Repository
git clone https://github.com/MongooseReborn/kalos-engine.git<br>cd kalos-engine
2. System Requirements
Linux OS (Ubuntu 22.04+ recommended)
NVIDIA GPU with CUDA Driver 12.0+ installed
Python 3.10+ (for telemetry monitor scripts)
3. Run Precompiled Executables
Interactive Terminal UI (TUI):<br>./bin/kalos_tui
Hardware Telemetry Profiler:<br>./bin/kalos_bench
Fast In-Memory CUDA Executor:<br>./bin/kalos_runner
4. Shared C/CUDA Libraries (./bin/)
libkalos_snn.so: Spiking Neural Cortex Engine (15.16 μs LIF Cortex)
libkalos_sam.so: Sparse Associative Memory Engine (465.12 μs Vector Recall)
libkalos_haptics.so: Physical Touch & FFT Audio Engine
libkalos_soul.so: Binary Cortical Persistence Format (.soul)
📄 Release Documentation & Whitepapers
docs/Whitepaper_Industrial.html: Master Industrial Release HTML Whitepaper (Amber / Crimson CRT Theme).
docs/WHITE_PAPER.md: Complete Master Architectural & Empirical Markdown Whitepaper.
docs/BENCHMARK_REPORT.md: Detailed Empirical Benchmark Matrix (7B to 72B Models).
docs/SETUP_GUIDE.md: Comprehensive Integration & Configuration Guide.
📜 License, Attribution & Contact
License: Licensed under the MIT License.
Authors: Mongoose & Kalos Engine Architecture Team @ BlackForest Studio (2026).
Contact & Inquiries: blackforest.team@proton.me
Hugging Face Hub: https://huggingface.co/MongooseReborn/kalos-engine
GitHub Repository: https://github.com/MongooseReborn/kalos-engine
About<br>Project Kalos: A Biological-Resonant C/CUDA Neuromorphic Engine Suite & Empirical Whitepapers v1.0<br>Resources<br>Readme<br>Activity<br>Stars<br>0 stars<br>Watchers<br>0 watching<br>Forks<br>0 forks<br>Report repository
Releases
Packages
Contributors
Languages
You can’t perform that action at this time.