Part 1: Why LLM Inference Frameworks Use torch.compile

Part 2: How Does PyTorch Compile Regular Python Code?

Part 3: How TorchDynamo Captures Python Execution

Part 4: Understanding FX Graphs in PyTorch