Facing errors running the liftsim baseline

[dineshresearch]

@TomorrowIsAnOtherDay I am unable to run the Liftsim baseline PARL/examples/LiftSim_baseline/DQN/ and facing the below error. Can any one please help me kindly.

Environment python 3.6.1 gym==0.18 paddlepaddle==1.8.5 cpu (as i could not find the cpu wheel for 1.5.1 and unable to compile it myself i have installed 1.8.5 wheel through pip) parl= (https://github.com/PaddlePaddle/PARL/tree/r1.3) branch rlschool==0.0.1

python demo.py [05-25 10:28:49 MainThread @logger.py:224] Argv: demo.py iterations: 100000000 Traceback (most recent call last): File "demo.py", line 48, in run_main(sys.argv[1:]) File "demo.py", line 41, in run_main dispatcher = RL_dispatcher(mansion_env, args.iterations) File "/content/PARL/examples/LiftSim_baseline/DQN/rl_benchmark/dispatcher.py", line 53, in init self._algorithm = DQN(self._model, hyperparas) File "/usr/local/lib/python3.6/dist-packages/parl/algorithms/fluid/dqn.py", line 39, in init assert isinstance(act_dim, int) AssertionError

[dineshresearch]

I have commented out that assert part but now i am facing the below error python demo.py iterations: 100000000 <rl_benchmark.model.RLDispatcherModel object at 0x7f1b5a1a50f0> {'action_dim': 22, 'lr': 0.0005, 'gamma': 0.998} [05-25 11:46:24 MainThread @machine_info.py:84] Cannot find available GPU devices, using CPU now. Traceback (most recent call last): File "demo.py", line 48, in run_main(sys.argv[1:]) File "demo.py", line 41, in run_main dispatcher = RL_dispatcher(mansion_env, args.iterations) File "/content/PARL/examples/LiftSim_baseline/DQN/rl_benchmark/dispatcher.py", line 56, in init self._act_dim) File "/content/PARL/examples/LiftSim_baseline/DQN/rl_benchmark/agent.py", line 34, in init super(ElevatorAgent, self).init(algorithm) File "/usr/local/lib/python3.6/dist-packages/parl/core/fluid/agent.py", line 80, in init self.build_program() File "/content/PARL/examples/LiftSim_baseline/DQN/rl_benchmark/agent.py", line 64, in build_program self._value = self.alg.define_predict(obs) AttributeError: 'DQN' object has no attribute 'define_predict'