Paddle cannot load a saved model file in Cambricon MLU370 Card

[wangzy0327]

bug描述 Describe the Bug

paddle version : PaddlePaddle:develop d2033d0

paddlepaddle == 2.6

device Cambricon MLU370 PaddleCustomDevice:mlu

env: docker registry.baidubce.com/device/paddle-mlu:ctr2.15.0-ubuntu20-x86_64-gcc84-py310

python script

import paddle
from paddle.static import InputSpec

def save_function():
     @paddle.jit.to_static
     def fun(inputs):
         return paddle.tanh(inputs)

     path = 'test_jit_save_load_function_1/func'
     inps = paddle.rand([3, 6])
     origin = fun(inps)

     paddle.jit.save(fun, path)
     load_func = paddle.jit.load(path)

     load_result = load_func(inps)
     print((load_result - origin).abs().max() < 1e-10)


print("save function ... load function...")
save_function()

bug report

I0228 09:59:05.639653 2263858 init.cc:234] ENV [CUSTOM_DEVICE_ROOT]=/usr/local/lib/python3.10/dist-packages/paddle_custom_device
I0228 09:59:05.639703 2263858 init.cc:143] Try loading custom device libs from: [/usr/local/lib/python3.10/dist-packages/paddle_custom_device]
I0228 09:59:05.670440 2263858 custom_device.cc:1099] Succeed in loading custom runtime in lib: /usr/local/lib/python3.10/dist-packages/paddle_custom_device/libpaddle-custom-mlu.so
I0228 09:59:05.677508 2263858 custom_kernel.cc:63] Succeed in loading 264 custom kernel(s) from loaded lib(s), will be used like native ones.
I0228 09:59:05.677863 2263858 init.cc:155] Finished in LoadCustomDevice with libs_path: [/usr/local/lib/python3.10/dist-packages/paddle_custom_device]
I0228 09:59:05.678017 2263858 init.cc:240] CustomDevice: mlu, visible devices count: 1
save function ... load function...
/usr/local/lib/python3.10/dist-packages/paddle/jit/api.py:898: UserWarning: What you save is a function, and `jit.save` will generate the name of the model file according to `path` you specify. When loading these files with `jit.load`, you get a `TranslatedLayer` whose inference result is the same as the inference result of the function you saved.
  warnings.warn(
/usr/local/lib/python3.10/dist-packages/paddle/static/io.py:618: UserWarning: no variable in your model, please ensure there are any variables in your model to save
  warnings.warn(
W0228 09:59:10.656574 2263858 tensor.cc:405] The `is_initialized` method is deprecated since version 2.3, and will be removed in version 2.4! Please use `initialized` method instead.
Traceback (most recent call last):
  File "/opt/paddle_tests/test_save_load.py", line 92, in <module>
    save_function()
  File "/opt/paddle_tests/test_save_load.py", line 87, in save_function
    load_result = load_func(inps)
  File "/usr/local/lib/python3.10/dist-packages/paddle/nn/layer/layers.py", line 1426, in __call__
    return self.forward(*inputs, **kwargs)
  File "/usr/local/lib/python3.10/dist-packages/paddle/jit/translated_layer.py", line 1475, in __i_m_p_l__
    return _run_dygraph(self, input, program_holder)
  File "/usr/local/lib/python3.10/dist-packages/paddle/jit/translated_layer.py", line 1004, in _run_dygraph
    _legacy_C_ops.run_program(
RuntimeError: (NotFound)  interface not found.
  [Hint: Expected dev_impl_map.find(device_type) != dev_impl_map.end(), but received dev_impl_map.find(device_type) == dev_impl_map.end().] (at /opt/Paddle/paddle/phi/backends/device_manager.cc:296)
  [operator < run_program > error]

其他补充信息 Additional Supplementary Information

No response

[wangzy0327]

@blacksheep-Aristotle Can you help me to solve the problem？