Welcome to your reinforcement learning assignment on AIcrowd!

Hi!

This is the welcome thread for your first RL assignment.

Welcome to the AIcrowd forum!

Here, you can post your questions, doubts, and any issues you face while making a submission or going through the notebooks and our team will get back to you soon. :running_man:‍♂

In the meanwhile, you can reply to this thread with a brief introduction of you and what excites you about Reinforcement learning. We’d love to hear your thoughts! :sparkles:


Check out some of our most popular Reinforcement learning challenges!

:steam_locomotive: Flatland Competition
:video_game: Procgen Competition
:gem: MineRL Competition

All the best!
Team AIcrowd

Hi!
I am Jaswanthi, a student crediting CS6700 course.


Could you kindly help with the following error that appears after submission.

Thank you.

Hi Jaswanthi,

The actions to be given for the policy in rl-taxi are strings, I see you submission is with integers. Please fix that.

Hi. I did that.
Still the same issue.

I see the list of states and possible actions is hardcoded in your code. Please remove all hardcoding. The list of possible states and possibles actions for each state are to be obtained from the environment.

env.possible_states
env.possible_actions(state)
env.ride_probabilities(state, action)

Thanks. That worked.

Hi dipam,
Its the same error (Inference failed) popping up with my VI submission.
Could you please let me know why this is happening.
I tried to go through the log files. But could not decode.
Please help me with this.

Thank you.

Hello

Here is the error traceback (excluding the color tags)

---------------------------------------------------------------------------
FileNotFoundError                         Traceback (most recent call last)
<ipython-input-1-365589ebf4ca> in <module>
      3 # # modify the default parameters of np.load
      4 np.load = lambda *a,**k: np_load_old(*a, allow_pickle=True, **k)
----> 5 data = np.load('/content/hw2_q2/inputs/env_params_0.npy')
      6 np.load = np_load_old
      7 env = GridEnv_HW2(data.item().get('goal_location'),

<ipython-input-1-365589ebf4ca> in <lambda>(*a, **k)
      2 np_load_old = np.load
      3 # # modify the default parameters of np.load
----> 4 np.load = lambda *a,**k: np_load_old(*a, allow_pickle=True, **k)
      5 data = np.load('/content/hw2_q2/inputs/env_params_0.npy')
      6 np.load = np_load_old

/opt/conda/lib/python3.8/site-packages/numpy/lib/npyio.py in load(file, mmap_mode, allow_pickle, fix_imports, encoding)
    414             own_fid = False
    415         else:
--> 416             fid = stack.enter_context(open(os_fspath(file), "rb"))
    417             own_fid = True
    418

FileNotFoundError: [Errno 2] No such file or directory: '/content/hw2_q2/inputs/env_params_0.npy'

Hi,
I removed it.
Yet I see the following error.

Hello @jaswanthi_mandalapu

There is a timeout of 900s for evaluation and the message above means that your code ran for more than 900s.

Hi,
I have modified accordingly and its working now.
Thanks for that.

But,
For both the problems my results matched to the target outputs provided.
Yet I see my scores to be 0.
Can you kindly let me know how I can go about this.

Thank you.

Hi jaswanthi_mandalapu,

Please provide submission ids for your 0 scoring submissions. But please check once before that if there is any hardcoding in your code.

Hi my submission ID is #126315
No. There is no hard coding. I have verified.

Thank you.

Hi jaswanthi_mandalapu,

The solutions to the input test cases are provided in the targets folder inside the hw2_q2 directory, please use those to check your results before submitting. You should get exactly equal policy and values that are nearly equal to upto 2-3 decimal points.

Note - Please don’t submit any code that tries to read from hw2_q2/targets, this folder is not available during test time.

In case you missed it, extra test cases are provided here

Hi Dipam,
Thanks for the clarification.
My code seems to work for the first 2 test cases provided.
The next three I have no idea where it is going wrong. My policies are correct but the cost values are not matching.
I shall work on it.
Thanks.

Hi,
For taxi_question, my code matches with all the target cases.
Yet I see the following inference failed error.
Could you please kindly help with this.


Thank you.

Hi @jaswanthi_mandalapu

The package “tabulate” is not installed, put the code for its installation on the top most cell marked for installation of packages.


Hi I am getting inference failed can you please why this issue is caused?
Thank You