-
Demo AI Quoting System with Fine-tuning and QuantizationA report + reflection on building demo AI quoting system by fine-tuning Qwen3-4B with a single RTX 4090, quantized to Q4_K_M GGUF, and benchmarked CPU vs GPU.
20 min read English -
Project Still: Private Digital Legacy StackA private digital-legacy prototype: real-time face/voice conversion and GLM-4-9B persona fine-tuning, shared at a high level to protect privacy.
15 min read English