Skip to content
View ChaosAIVision's full-sized avatar

Block or report ChaosAIVision

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 supported. This note will be visible to only you.
Report abuse

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

Report abuse
.github/profile/README.md

Hi there, I'm Chaos!

Profile views

AI Engineer | Generative AI & Computer Vision | Ho Chi Minh City


About Me

Coding

  • I'm currently working on AI Agent Systems, RAG Integration, and Model Optimization
  • I'm currently learning Advanced inference optimization & distributed training
  • I'm looking to collaborate on open source ML/AI projects
  • Ask me about LLM fine-tuning, diffusion models, computer vision, and ML deployment
  • How to reach me: [email protected]

Featured Projects

TrainForge

Easy LLM/VLM Training Framework

  • Framework for LoRA, quantization (4/8-bit), multi-GPU fine-tuning
  • Fine-tuned Qwen3 4B for RAG reasoning
  • 0.9894 context_recall | 0.825 faithfulness (RAGAS)
  • Minimum 8GB GPU required
View Repository

Light-Diffusion

VRAM-Efficient Diffusion Framework

  • Object insertion & image editing pipeline
  • Stable Diffusion Inpainting 1.5 optimization
  • Trains on minimum 8GB GPU
View Repository

YOLO-AI Framework

Easy Deploy an Detection System with YOLO

  • YOLOv8 + ONNX + BentoML full pipeline
  • React-based UI for image/video streaming
  • Achieves 18-20 FPS performance
View Repository

Current Focus

Active Development

  • Multi-agent chatbots with LangGraph
  • Advanced RAG pipelines & agent systems
  • Model evaluation & benchmarking tools
  • MLOps automation utilities

Languages and Tools

Programming Languages

Python

AI/ML & Deep Learning

PyTorch Hugging Face

Model Training & Optimization

PyTorch Lightning Unsloth bitsandbytes PEFT

Inference & Deployment

VLLM TensorRT Triton ONNX OpenVINO BentoML

Agent Systems & Frameworks

LangGraph LangChain

Databases & Storage

PostgreSQL SQLite Qdrant


Work Experience

Working

AI Engineer | Pythera AI (Aug 2024 - Oct 2025)

  • Designed and implemented multi-agent chatbot systems using LangGraph and custom MCP tools
  • Built RAG-based workflows with structured data retrieval and reasoning
  • Developed workflow orchestration with Windmill for agent communication
  • Fine-tuned DeepSeek-R1 8B on 16GB GPU with LoRA optimization
  • Deployed models via VLLM achieving 59 TPS inference on RTX 5090
  • Converted Stable Diffusion Inpainting to TensorRT + Triton (6s/image generation)

AI Engineer Intern | QAI – FPT Software (Jan 2024 - Apr 2024)

  • Collected and preprocessed 20k+ PPE images from industrial environments
  • Trained YOLOv8 detection model achieving 0.95 mAP@50

Current Focus

  • Improving model evaluation and benchmarking workflows
  • Building lightweight utilities for rapid ML prototyping
  • Exploring advanced inference optimization techniques
  • Contributing to open-source ML communities
  • Multi-agent systems & LLM applications

Connect with Me

LinkedIn GitHub HuggingFace Email


@ChaosAIVision's activity is private