我通过Beta类中的_sample（）函数采样一个样本，并用_log_prob()函数求其对数概率，但程序报错显示gamma算子不支持在CPU使用

源代码如下

class Actor_Beta(nn.Cell):

def get_dist(self, s):
    alpha, beta = self.forward(s)
    dist = nn.probability.distribution.Beta(alpha, beta)
    return dist

self.actor = Actor_Beta(args)

dist = self.actor.get_dist(s)

a = dist._sample  
a_logprob = dist._log_prob(a)

报错显示如下

Traceback (most recent call last): 

  File "E:\test\pythonProject2\train\sim_train.py", line 18, in <module> 

    main(args) 

  File "E:\test\pythonProject2\train\sim_train.py", line 10, in main 

    runner.run(60 * 10)  # 10min 

  File "E:\test\pythonProject2\runner\shared\sim_runner.py", line 111, in run 

    a, a_logprob = agent.choose_action(s)  # Action and the corresponding log probability 

  File "E:\test\pythonProject2\algorithms_ms\algorithm_ms\actor_critic_ms.py", line 173, in choose_action 

    a_logprob = dist._log_prob(a)  # The log probability density of the action 

  File "E:\test\pythonProject2\.venv\lib\site-packages\mindspore\nn\probability\distribution\beta.py", line 323, in _log_prob 

    return log_unnormalized_prob - self.lbeta(concentration1, concentration0) 

  File "E:\test\pythonProject2\.venv\lib\site-packages\mindspore\nn\cell.py", line 705, in __call__ 

    raise err 

  File "E:\test\pythonProject2\.venv\lib\site-packages\mindspore\nn\cell.py", line 701, in __call__ 

    output = self._run_construct(args, kwargs) 

  File "E:\test\pythonProject2\.venv\lib\site-packages\mindspore\nn\cell.py", line 482, in _run_construct 

    output = self.construct(*cast_inputs, **kwargs) 

  File "E:\test\pythonProject2\.venv\lib\site-packages\mindspore\nn\layer\math.py", line 735, in construct 

    log_beta_one_large = self.lgamma(x_min) + log_gamma_difference_big_y 

  File "E:\test\pythonProject2\.venv\lib\site-packages\mindspore\nn\cell.py", line 705, in __call__ 

    raise err 

  File "E:\test\pythonProject2\.venv\lib\site-packages\mindspore\nn\cell.py", line 701, in __call__ 

    output = self._run_construct(args, kwargs) 

  File "E:\test\pythonProject2\.venv\lib\site-packages\mindspore\nn\cell.py", line 482, in _run_construct 

    output = self.construct(*cast_inputs, **kwargs) 

  File "E:\test\pythonProject2\.venv\lib\site-packages\mindspore\nn\layer\math.py", line 237, in construct 

    input_dtype = self.dtype(x) 

  File "E:\test\pythonProject2\.venv\lib\site-packages\mindspore\ops\operations\array_ops.py", line 268, in __call__ 

    return x.dtype 

  File "E:\test\pythonProject2\.venv\lib\site-packages\mindspore\common\_stub_tensor.py", line 94, in dtype 

    self.stub_dtype = self.stub.get_dtype() 

RuntimeError:  

---------------------------------------------------- 

- Kernel select failed: 

---------------------------------------------------- 

Unsupported op [Gamma] on CPU, Please confirm whether the device target setting is correct, or refer to 'mindspore.ops' at https://www.mindspore.cn to query the operator support list. 

node: @kernel_graph_14:CNode_4{[0]: ValueNode<Primitive> Gamma, [1]: ValueNode<Tensor> Tensor(shape=[2], dtype=Int64, value=[1 1]), [2]: @kernel_graph_14:param_Parameter_5, [3]: @kernel_graph_14:param_Parameter_6} 

---------------------------------------------------- 

- C++ Call Stack: (For framework developers) 

---------------------------------------------------- 

mindspore\ccsrc\plugin\device\cpu\hal\hardware\cpu_device_context.cc:386 mindspore::device::cpu::CPUKernelExecutor::SetOperatorInfo