Overview Building and using a TensorRT engine for inference involves two distinct phases: build time (optimization) and runtime (execution). Build time (optimization and compilation) Load model typically from format like ONNX. Parse model, understand network structure and weights. Optimize model, including layer, …

Read More

Fundamentals of Reinforcement Learning
Read More

Based on NG's Machine Learning course
Read More

Probability Knowledge Based on Deep Learning (2017, MIT) book.
Read More