lerrel / rllab-adv Goto Github PK

When running this code, there is a problem showed in the title. I use my MacBook to run it with the newest operating system version, how to fix it? Thanks.

Hi,
I am trying to reproduce the code but I am getting the following error :

/home/mujtaba.hasan/anaconda3/envs/rllab3-adv/lib/python3.5/site-packages/theano/tensor/signal/downsample.py:6: UserWarning: downsample module has been moved to the theano.tensor.signal.pool module.
"downsample module has been moved to the theano.tensor.signal.pool module.")
2018-06-07 16:53:29.812020 IST | Warning: skipping Gym environment monitoring since snapshot_dir not configured.
Traceback (most recent call last):
File "/home/mujtaba.hasan/anaconda3/envs/rllab3-adv/lib/python3.5/site-packages/gym/envs/registration.py", line 143, in spec
return self.env_specs[id]
KeyError: 'InvertedPendulumAdv-v1'

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "adversarial/scripts/train_adversary.py", line 73, in
env = normalize(GymEnv(env_name, adv_fraction))
File "/home/mujtaba.hasan/Documents/RL/rllab-adv/rllab/envs/gym_env.py", line 54, in init
env = gym.envs.make(env_name)
File "/home/mujtaba.hasan/anaconda3/envs/rllab3-adv/lib/python3.5/site-packages/gym/envs/registration.py", line 167, in make
return registry.make(id)
File "/home/mujtaba.hasan/anaconda3/envs/rllab3-adv/lib/python3.5/site-packages/gym/envs/registration.py", line 118, in make
spec = self.spec(id)
File "/home/mujtaba.hasan/anaconda3/envs/rllab3-adv/lib/python3.5/site-packages/gym/envs/registration.py", line 153, in spec
raise error.UnregisteredEnv('No registered env with id: {}'.format(id))
gym.error.UnregisteredEnv: No registered env with id: InvertedPendulumAdv-v1

I got the following error when trying to run:
python adversarial/scripts/train_adversary.py --env InvertedPendulumAdv-v1 --folder ~/rllab-adv/results

The error message:

Traceback (most recent call last): File "adversarial/scripts/train_adversary.py", line 161, in <module> args=args File "/home/zhsu/DRL/rllab-adv2/rllab-adv/rllab/policies/gaussian_mlp_policy.py", line 110, in __init__ mean_var, log_std_var = L.get_output([l_mean, l_log_std]) File "/home/zhsu/DRL/env/lib/python3.5/site-packages/lasagne/layers/helper.py", line 185, in get_output all_outputs[layer] = layer.get_output_for(layer_inputs, **kwargs) File "/home/zhsu/DRL/rllab-adv2/rllab-adv/rllab/core/lasagne_layers.py", line 30, in get_output_for tiled = TT.tile(reshaped_param, tile_arg, ndim=ndim) File "/home/zhsu/DRL/env/lib/python3.5/site-packages/theano/tensor/basic.py", line 4398, in tile for r in reps]): File "/home/zhsu/DRL/env/lib/python3.5/site-packages/theano/tensor/basic.py", line 4395, in <listcomp> if not numpy.all([isinstance(r, integer_types) or File "/home/zhsu/DRL/env/lib/python3.5/site-packages/theano/tensor/var.py", line 514, in __iter__ for i in xrange(theano.tensor.basic.get_vector_length(self)): File "/home/zhsu/DRL/env/lib/python3.5/site-packages/theano/tensor/basic.py", line 3907, in get_vector_length raise ValueError("length not known") ValueError: length not known

Hello,

I'm running into a lot of issues trying to get your code up and running. I have tried conda, which failed following which I tried my own enviroment manually picking versions most likely to work with your code.

Can you please specify the exact versions/commit shas of what you're using for the following software?

• rllab
• mujoco-py
• gym
• mjpro

I was trying to resolve the errors one by one, the last one I landed on is:

Traceback (most recent call last):
  File "adversarial/scripts/train_adversary.py", line 73, in <module>
    env = normalize(GymEnv(env_name, adv_fraction))
  File "/opt/rarl/env/lib/python3.5/site-packages/rllab-0.1.0-py3.5.egg/rllab/envs/gym_env.py", line 55, in __init__
    def_adv = env.adv_action_space.high[0]
AttributeError: 'HopperEnv' object has no attribute 'adv_action_space'