TinyTTS: 1.6M-parameter TTS runs at 53x real-time on CPU

Developer tronghieuit has released TinyTTS, an ultra-lightweight English text-to-speech model that generates natural speech with only about 1.6M parameters, publishing it on GitHub. At roughly 3.4MB in ONNX (FP16) format, the model is extremely compact and delivers speech synthesis well above real time on a CPU alone, without a GPU.