4 ms·Low-Latency Bayesian Inference: Deploying Models with PyTorch and ONNX7 points by apetrov 2y agocontrolweather 2y ago[dead]