Jin-Ho Kwon

ML Engineer in Seoul. PyTorch, ONNX, and model compression. Background in on-device AI for consumer electronics. The model that ships is never the model that won the benchmark