Error When Training Agent

[stehrbaer]

When training the agent using the FinRL, I keep getting the same error that the tensor position must be a tuple not map, and not quite sure how to address the issue?

I am using a Google Collab on Mac. ERROR:

TypeError Traceback (most recent call last) in 13 erl_params=ERL_PARAMS, 14 cwd='./papertrading_erl', #current_working_dir ---> 15 break_step=1e5)

4 frames /usr/local/lib/python3.7/dist-packages/elegantrl/agents/AgentBase.py in convert_trajectory(failed resolving arguments) 286 287 '''stack items''' --> 288 traj_state = torch.stack(traj_list1[0]) 289 traj_action = torch.stack(traj_list1[3]) 290

TypeError: stack(): argument 'tensors' (position 1) must be tuple of Tensors, not map

[quantumiracle]

This problem can be fixed at the ElegantRL side, by change this line to be:

        traj_list1 = list(map(list, zip(*traj_list)))  # state, reward, done, action, noise

However, there are still some other bugs at the ElegantRL side.