Skip to content
View Yog-Sotho's full-sized avatar
💭
Chillin' and vibe-coding 🤖
💭
Chillin' and vibe-coding 🤖

Block or report Yog-Sotho

Report abuse

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

Report abuse
Yog-Sotho/README.md

🜏 Yog-Sotho

Exploring intelligence, one experiment at a time.


🧠 About

  • Building and experimenting with LLMs and intelligent systems
  • Focus on fine-tuning, evaluation, and control
  • Creating tools that turn models into usable systems
  • Experimenting with security and stuff

🧪 Live Experiment Dashboard

Auto-updating signals from ongoing work


🧬 Research Logs

🧠 LLM Fine-Tuner

  • Experimenting with model adaptation pipelines
  • Interface-driven training workflows (Gradio)
  • Focus: usability + control of model behavior

🧩 Skills / Modular AI Systems

  • Designing composable “skills” for LLMs
  • Exploring structured prompting + tool use
  • Goal: reusable intelligence components

🌑 Nocturna Trading System

  • Experimental project (ML / crypto trading)
  • Focus on eliminate the risks of monolithic execution, such as race conditions, blocking I/O, and state loss during crashes.

⚙️ Tech

Languages: Python, JavaScript
Focus:     LLMs, APIs, Tooling, Experimentation
Stack:     Gradio, Model APIs, Data Pipelines

🌐 Links


🜏 Support

If you like my work, feel free to send me a tip: 🪙 BTC (BSC network) 0x4b038d095ccb5bd088b8cdc33d4a202372647f2c 💎 ETH (ETH network) 0x4b038d095ccb5bd088b8cdc33d4a202372647f2c 🚧 BNB (BSC network) 0x4b038d095ccb5bd088b8cdc33d4a202372647f2c

Pinned Loading

  1. LLM-fine-tuner LLM-fine-tuner Public

    Powerful no-code LLM fine-tuner: upload data → train → deploy in minutes. Unsloth 2-5× acceleration · QLoRA/DPO/RLHF/PPO/ORPO · Reward Model training · GGUF export · vLLM inference · BLEU/ROUGE/BER…

    Python 32 4

  2. Brainbrew Brainbrew Public

    A simple GUI tool that generates LLM training datasets through model distillation. Designed for **non-technical users**.

    Python 7 1

  3. Nocturna-Trading-System Nocturna-Trading-System Public

    NOCTURNA v2.0 - Institutional-Grade Algorithmic Trading System ## 🏛 System Architecture: Distributed Async Model NOCTURNA v2.0 is a distributed, event-driven trading system. It is engineered to eli…

    Python 5

  4. LiSa-LLM LiSa-LLM Public

    a tiny, sandbox‑run LLM client without any llama.cpp dependency. It uses only the GGML library and implements the transformer forward pass itself.

    C++ 1

  5. github-mcp-server-claudeai github-mcp-server-claudeai Public

    A simple MCP server to allow Claude.ai to connect to Github

    TypeScript

  6. Ollama-Remote-Server-Scanner Ollama-Remote-Server-Scanner Public

    This tool scans IP ranges for open LLM servers (Ollama, LM Studio, TextGen WebUI) and reports available models, loaded processes, and model configurations.

    Python 1