Torch compile caching for inference speed

In three linesReplicate implements PyTorch torch.compile caching to reduce boot and inference times. Compiled models are cached across invocations, eliminating recompilation on each run.

Summary generated by Claude — human-verified