I think it’s safe to assume that the Inference API isn’t working properly at the moment, including Qwen 32B Coder…