Skip to content
View loki-smip's full-sized avatar
2 followers · 1 following

Achievements

Achievement: YOLO

Achievements

Achievement: YOLO

Block or report loki-smip

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
loki-smip/README.md

👋 Hi, I'm @loki-smip

🚀 Recursive AI Architect & Systems Administrator

I am a developer and sysadmin focused on efficient LLM architectures and high-performance self-hosted infrastructure. I bridge the gap between deep learning and bare-metal systems.

🛠️ Technical Arsenal

  • Languages: Rust | C++ | C | Python | Java
  • AI/ML: Recursive HRM, MLA (Multi-Head Latent Attention), SwiGLU, Custom Tokenizers
  • Infrastructure: Home Server Sysadmin, Docker, Local Intranet Management
  • Hardware: Diving deep into GPU/CPU architecture and AI compute optimization.

🧬 Featured Project: Ouroboros-Recursive-LM

  • Architecture: A Recursive Hierarchical Representation Model (R-HRM).
  • Innovation: Uses 8 recursive steps to maximize reasoning depth in a compact 512-D space.
  • Goal: Hitting the 2.1 Loss "Sniper Target" for optimal reasoning convergence.

🎨 Creative & Tech Content

I document my experiments in technology and digital media.

🌱 Current Focus

  • 🌱 Learning: Advanced AI Training, GPU/CPU hardware internals, and Docker orchestration.
  • 💞️ Collaboration: Efficient LLM inference engines and high-performance computing.
  • 📫 Reach me: Open an Issue
  • ⚡ Fun Fact: I run my own intranet and train recursive brains.

Popular repositories Loading

  1. mouse mouse Public

    HTML 1

  2. Ouroboros-Recursive-LM Ouroboros-Recursive-LM Public

    Architecture: Explains Recursive Blocks, SwiGLU, and MLA. Usage: How to run train.py (CPU vs GPU) and chat.py. Data: Credits the datasets (TinyGSM, Cosmopedia, etc.) and explains the "Clean Filter".

    Python 1

  3. esp-mic esp-mic Public

    Stream audio from an INMP441 I2S microphone over Wi-Fi using an ESP8266, then receive and transcribe it on your PC in real-time using Python.

    Python 1

  4. EvoTransformer EvoTransformer Public

    a ai model that self-correction loop and self evolves its self and is super efficient and can run on cpu and can be trained on cpu or gpu

    Python 1

  5. openbot-pc-server-projuct- openbot-pc-server-projuct- Public

    Autonomous RC car using ESP8266 + ESP32-CAM. Drive it, record data, train a CNN, let it drive itself — all from a browser dashboard.

    Python 1

  6. projust projust Public

    C++