[Core] draft_model_runner: Implement prepare_inputs on GPU for advance_step (#6338)

csrc/torch_bindings.cpp

@@ -72,6 +72,10 @@ TORCH_LIBRARY_EXPAND(TORCH_EXTENSION_NAME, ops) {
   ops.def("gelu_quick(Tensor! out, Tensor input) -> ()");
   ops.impl("gelu_quick", torch::kCUDA, &gelu_quick);
 
+  // prepare_inputs advance_step
+  ops.def("advance_step", &advance_step);
+  ops.impl("advance_step", torch::kCUDA, &advance_step);
+
   // Layernorm
   // Apply Root Mean Square (RMS) Normalization to the input tensor.
   ops.def(