Auto-scheduler layout transform

I executed the following case:

x_shape = (1, 3, 6, 6)
x_np = np.random.random_sample(x_shape).astype("float32")
x = relay.var("x", shape=x_shape, dtype=dtype)
y = relay.op._make.auto_scheduler_layout_transform(x, "NCHW", "NHWC")
func = relay.Function([x], y)
inputs = {"x":tvm.nd.array(x_np)}
mod = tvm.IRModule().from_expr(func)
params = {}
with tvm.transform.PassContext(opt_level=2):
   exe = relay.backend.vm.compile(mod=mod, params=params, target="llvm")
vm = tvm.runtime.vm.VirtualMachine(exe, device=tvm.device("llvm"))
out = vm.run(**{**params, **inputs})

But it raised an error:

File "/tvm/include/tvm/topi/transform.h", line 1585
        at /tvm/include/tvm/topi/transform.h:1585
TVMError: Check failed: dst_indices_expr.size() == dst_axes.size() (1 vs. 2) :

Could you help me solve this problem?