Expected all tensors to be on the same device, but found at least two devices, cpu and cuda:0! (when checking argument for argument mat1 in method wrapper_CUDA_addmm)

RuntimeError Traceback (most recent call last) Cell In[25], line 3 1 from elegantrl.train.run import train_agent ----> 3 train_agent(args)

File E:\langs\anaconda3\envs\deep-fin\lib\site-packages\elegantrl\train\run.py:89, in train_agent(failed resolving arguments) 86 buffer[:] = buffer_items 88 torch.set_grad_enabled(True) ---> 89 logging_tuple = agent.update_net(buffer) 90 torch.set_grad_enabled(False) 92 evaluator.evaluate_and_save(actor=agent.act, steps=horizon_len, exp_r=exp_r, logging_tuple=logging_tuple)

File E:\langs\anaconda3\envs\deep-fin\lib\site-packages\elegantrl\agents\AgentSAC.py:29, in AgentSAC.update_net(self, buffer) 25 with torch.no_grad(): 26 states, actions, rewards, undones = buffer.add_item 27 self.update_avg_std_for_normalization( 28 states=states.reshape((-1, self.state_dim)), ---> 29 returns=self.get_cumulative_rewards(rewards=rewards, undones=undones).reshape((-1,)) 30 ) 32 '''update network''' 33 obj_critics = 0.0

File E:\langs\anaconda3\envs\deep-fin\lib\site-packages\elegantrl\agents\AgentBase.py:190, in AgentBase.get_cumulative_rewards(self, rewards, undones) 188 last_state = self.last_state 189 next_action = self.act_target(last_state) --> 190 next_value = self.cri_target(last_state, next_action).detach() 191 for t in range(horizon_len - 1, -1, -1): 192 returns[t] = next_value = rewards[t] + masks[t] * next_value

File E:\langs\anaconda3\envs\deep-fin\lib\site-packages\torch\nn\modules\module.py:1553, in Module._wrapped_call_impl(self, *args, kwargs) 1551 return self._compiled_call_impl(*args, *kwargs) # type: ignore[misc] 1552 else: -> 1553 return self._call_impl(args, kwargs)

File E:\langs\anaconda3\envs\deep-fin\lib\site-packages\torch\nn\modules\module.py:1562, in Module._call_impl(self, *args, *kwargs) 1557 # If we don't have any hooks, we want to skip the rest of the logic in 1558 # this function, and just call forward. 1559 if not (self._backward_hooks or self._backward_pre_hooks or self._forward_hooks or self._forward_pre_hooks 1560 or _global_backward_pre_hooks or _global_backward_hooks 1561 or _global_forward_hooks or _global_forward_pre_hooks): -> 1562 return forward_call(args, **kwargs) 1564 try: 1565 result = None

File E:\langs\anaconda3\envs\deep-fin\lib\site-packages\elegantrl\agents\net.py:415, in CriticEnsemble.forward(self, state, action) 414 def forward(self, state, action): --> 415 values = self.get_q_values(state=state, action=action) 416 value = values.mean(dim=1, keepdim=True) 417 return value

File E:\langs\anaconda3\envs\deep-fin\lib\site-packages\elegantrl\agents\net.py:422, in CriticEnsemble.get_q_values(self, state, action) 420 state = self.state_norm(state) 421 sa_tmp = self.encoder_sa(torch.cat((state, action), dim=1)) --> 422 values = torch.concat([dec_q(sa_tmp) for dec_q in self.decoder_qs], dim=1) 423 values = self.value_re_norm(values) 424 return values

File E:\langs\anaconda3\envs\deep-fin\lib\site-packages\elegantrl\agents\net.py:422, in (.0) 420 state = self.state_norm(state) 421 sa_tmp = self.encoder_sa(torch.cat((state, action), dim=1)) --> 422 values = torch.concat([dec_q(sa_tmp) for dec_q in self.decoder_qs], dim=1) 423 values = self.value_re_norm(values) 424 return values

File E:\langs\anaconda3\envs\deep-fin\lib\site-packages\torch\nn\modules\container.py:219, in Sequential.forward(self, input) 217 def forward(self, input): 218 for module in self: --> 219 input = module(input) 220 return input

File E:\langs\anaconda3\envs\deep-fin\lib\site-packages\torch\nn\modules\linear.py:117, in Linear.forward(self, input) 116 def forward(self, input: Tensor) -> Tensor: --> 117 return F.linear(input, self.weight, self.bias)

RuntimeError: Expected all tensors to be on the same device, but found at least two devices, cpu and cuda:0! (when checking argument for argument mat1 in method wrapper_CUDA_addmm)

运行的是quickstart_Pendulum_v1.ipynb

AI4Finance-Foundation / ElegantRL

Expected all tensors to be on the same device, but found at least two devices, cpu and cuda:0! (when checking argument for argument mat1 in method wrapper_CUDA_addmm) #369