Python VMM API¶
KV-cache virtual memory is used through the C++ VMM API and the PyTorch
torch.ops.lpu.vmm_* operators. Both APIs manage the
HW pages; from Python, call the PyTorch operators.
KV-cache virtual memory is used through the C++ VMM API and the PyTorch
torch.ops.lpu.vmm_* operators. Both APIs manage the
HW pages; from Python, call the PyTorch operators.