4 ms·

GLM-5.2 is a huge model. I don't think it would fit on the AI HAT+ 2 even if you quantized it to 2 bits