(NotFound) There are no kernels which are registered in the einsum operator. about quantum

[LeiZhang-116-4]

This bug has been fixed by the latest version of PaddleQuantum.
Note that token is required to access QPU. See here for more details.

[buzzgamesnews]

also encountered this bug
Traceback (most recent call last): File "main.py", line 79, in <module> main() File "main.py", line 65, in main cut_bitstring, _ = find_cut(G, p, ITR, LR, print_loss=True, plot=True) File "c:\dropbox\python\quantum\quantum\paddle_quantum\QAOA\maxcut.py", line 87, in find_cut loss = -loss_func(state) File "C:\Users\user\anaconda3\envs\qnet_env\lib\site-packages\paddle\fluid\dygraph\layers.py", line 930, in __call__ return self._dygraph_call_func(*inputs, **kwargs) File "C:\Users\user\anaconda3\envs\qnet_env\lib\site-packages\paddle\fluid\dygraph\layers.py", line 915, in _dygraph_call_func outputs = self.forward(*inputs, **kwargs) File "c:\dropbox\python\quantum\quantum\paddle_quantum\loss\measure.py", line 117, in forward output_state = paddle.einsum('ia, ab->ib', matrix, _state_data).reshape([2 ** num_qubits]) File "C:\Users\user\anaconda3\envs\qnet_env\lib\site-packages\paddle\tensor\einsum.py", line 993, in einsum return einsum_v2(equation, *operands) File "C:\Users\user\anaconda3\envs\qnet_env\lib\site-packages\paddle\tensor\einsum.py", line 776, in einsum_v2 return gen_einsum_op(lhs + '->' + rhs, *operands) File "C:\Users\user\anaconda3\envs\qnet_env\lib\site-packages\paddle\tensor\einsum.py", line 805, in gen_einsum_op return _C_ops.einsum(operands, RuntimeError: (NotFound) There are no kernels which are registered in the einsum operator. [Hint: Expected kernels_iter != all_op_kernels.end(), but received kernels_iter == all_op_kernels.end().] (at C:\home\workspace\Paddle_release\paddle\fluid\imperative\prepared_operator.cc:341) [operator < einsum > error]

this exsample:
cd paddle_quantum/QAOA/example
python main.py

and TSP problem error in loss_func

Define the loss function

def loss_func(cir: Circuit, H: Hamiltonian) -> paddle.Tensor:
state = cir()
loss = paddle_quantum.loss.ExpecVal(H)
return loss(state) <-- HERE

[LeiZhang-116-4]

The function paddle.einsum in PaddlePaddle 2.3.1 does not work properly for complex tensors. Thus the problem should be solved by downgrading your PaddlePaddle version to 2.3.0.

[buzzgamesnews]

on default installed 2.3.2
when I try install PaddlePaddle 2.3.0
python -m pip install PaddlePaddle 2.3.0
got error:
Collecting PaddlePaddle
Using cached paddlepaddle-2.3.2-cp39-cp39-win_amd64.whl (64.3 MB)
ERROR: Could not find a version that satisfies the requirement 2.3.0 (from versions: none)
ERROR: No matching distribution found for 2.3.0

How install 2.3.0 ?

[buzzgamesnews]

Install it with - python -m pip install PaddlePaddle 2.3.0
fork fine!
thx

[buzzgamesnews]

but. how to connect to real quantum computer for comparision performance for TSP example ?
I try set backend:
paddle_quantum.set_backend('quleaf')

and got error :
Traceback (most recent call last):
File "C:\Dropbox\Python\Quantum\tsp.py", line 98, in
loss = loss_func(cir, H_C)
File "C:\Dropbox\Python\Quantum\tsp.py", line 61, in loss_func
return loss(state)
File "C:\Program Files\Python39\lib\site-packages\paddle\fluid\dygraph\layers.py", line 930, in call
return self._dygraph_call_func(*inputs, **kwargs)
File "C:\Program Files\Python39\lib\site-packages\paddle\fluid\dygraph\layers.py", line 915, in _dygraph_call_func
outputs = self.forward(*inputs, **kwargs)
File "C:\Program Files\Python39\lib\site-packages\paddle_quantum\loss\measure.py", line 94, in forward
paddle.concat(state.param_list),
File "C:\Program Files\Python39\lib\site-packages\paddle\tensor\manipulation.py", line 331, in concat
return paddle.fluid.layers.concat(input=x, axis=axis, name=name)
File "C:\Program Files\Python39\lib\site-packages\paddle\fluid\layers\tensor.py", line 343, in concat
_C_ops.concat(input, out, 'axis', axis)
ValueError: (InvalidArgument) concat(): argument 'X' (position 0) must be list of Tensors, but got empty list (at C:\home\workspace\Paddle_release\paddle\fluid\pybind\op_function_common.cc:768)

[wangzihe1996]

This is a bug. Briefly, it will occur when your circuit has no trainable parameters.

[buzzgamesnews]

without backend bug not appears.
but the backend set leads to failure TSP example:
paddle_quantum.set_backend('quleaf')

[wangzihe1996]

It only occurs when the backend is quleaf.

[buzzgamesnews]

please advise correct backend ?

[wangzihe1996]

If you want to use the simulator, you can choose the 'state_vector' or 'density_matirx'. If you want to use the real quantum computer, you can just choose the 'quleaf' backend.

[buzzgamesnews]

hm. I do that, but not work. tutorials not work with real quleaf backend ? only simulations ?