Yugo Ito@yugorou010420066
2
Events attended
2
Submissions made
Japan
2+ years of experience
About me
High school researcher from Japan working on imitation learning for robotics (GYOZA, SO-101 arm) and Japanese LLM alignment. AMD Developer Hackathon HF Community Prize winner. Open-source contributor (mlx-lm), creator of Shiin IME and TeenEmo.
🤝 Top Collaborators
.png&w=640&q=75)
🤓 Latest Submissions
YOUSUN Axion
YOUSUN Axion is an AI-powered developer platform built to help teams prepare AI workloads for AMD GPU environments. Many AI projects are still written with CUDA-specific assumptions, heavy FP32 memory usage, unclear benchmark results, and limited ROCm readiness. YOUSUN Axion solves this by giving developers a complete workflow instead of only generic suggestions. The platform scans AI code or error logs, detects CUDA/ROCm compatibility issues, identifies memory and precision problems, and gives an AMD readiness score. It then acts like an autonomous GPU engineer by diagnosing the root cause, recommending practical optimizations, generating code-level patch suggestions, and comparing before/after performance. YOUSUN Axion focuses on real developer needs: reducing debugging time, improving deployment confidence, and making AMD GPU migration easier. It can recommend changes like using BF16 precision, reducing batch size, replacing hardcoded CUDA usage, enabling inference mode, and preparing workloads for ROCm-based execution. The project also includes a benchmark and reporting layer. It can show latency, throughput, memory usage, readiness improvements, and generate a professional AMD Readiness Report that developers can share with teams, clients, or judges. In simple terms, YOUSUN Axion helps developers move from broken, slow, or non-optimized AI workloads to AMD-ready, optimized, benchmarked deployments. It is designed for AI developers, ML engineers, MLOps teams, and builders who want a clearer path to running AI applications on AMD infrastructure.
13 Jul 2026
Lumi, AI voice helper for elderly with dementia
Lumi is a domain fine-tuned AI voice companion for dementia and Alzheimer's care. Built on AMD MI300X using QLoRA and GRPO reinforcement learning, Lumi handles confusion, repetition, and emotional fragility the way a trained caregiver would — not a generic chatbot. 55 million people live with dementia worldwide. Families cannot provide 24/7 care — and existing AI companions fail them. They reset every session, correct temporal confusion (which is clinically harmful), and leave patients vulnerable to scams costing $3 billion annually in elder fraud. Lumi is the first AI companion purpose-built for this population. Fine-tuned on AMD MI300X using QLoRA and GRPO reinforcement learning — the same technique behind DeepSeek-R1 — Lumi was trained on 8,540 dementia-specific samples processed through our EQ-Matrix framework, covering scenarios across severity levels, emotional states, and scam patterns. Persistent memory via ChromaDB injects prior session context into every new conversation — patients never repeat themselves. A structured output format fires the opening spoken line to TTS before the full response is generated, achieving time-to-first-audio under 1.5 seconds. A binary scam deflection classifier intercepts fraud attempts gently, without alarming the patient. On EQ-Bench 3, Lumi ranked 7th out of 46 models with a Rubric Score of 14.55 — confirming genuine emotional intelligence gains, not just surface fluency. The entire pipeline runs locally on AMD MI300X with ROCm and vLLM. No data leaves the device. No proprietary APIs. Fully private, fully open hardware.
10 May 2026
