why there is AssertionError happened in MARL_MOUDLE/envs/tasks when i run this project?

[GengYiran]

Can you provide a screenshot of the error report, or provide the specific content of the error report?

[Psycho-9]

the error when i run the ## Close Door ### State experiment is as follows Simulator: number of environments 352 Traceback (most recent call last): File "train.py", line 69, in train() File "train.py", line 48, in train task, env = parse_task(args, cfg, cfg_train, sim_params, agent_index, logdir) File "/home/pyf/RLAfford-main/MARL_Module/envs/utils/parse_task.py", line 80, in parse_task task = eval(args.task)( File "/home/pyf/RLAfford-main/MARL_Module/envs/tasks/franka_cabinet.py", line 57, in init assert(self.env_num_train % self.cabinet_num_train == 0) AssertionError Printing Profile: End of Profile

[boshi-an]

For simplicity for implement, every object should be replicated for the same number. Which means # of different objects (cabinet_num_train) should be a factor of # of total environments. You can assign proper numbers in the command line when running the code using --num_xxx=xxx.

[Psycho-9]

how can i know the proper number of each experiments?

[GengYiran]

First, you can modify the number of objects or object types through the parameters in the python command or the config of the task, just make sure that the number of objects is a multiple of the object types (e.g., 4 objects of 2 types). Second, our experiments generally use the most objects to train, you can refer to the number of objects listed in the config is the upper limit of object types (of course, you can also add your own object dataset).

[GengYiran]