Any questions about the assignment? Ask them here!


If you are facing any issues with the notebook or with the AIcrowd platform, will free to ask your questions here or in a private message to me :slight_smile:

We will make sure that your queries are resolved :rocket:


The leaderboard for Value Iteration is filled with mostly zeros. Is there any bug ?

Hi, the env params loaded in the notebook dont seem to correlate to the grid world 1 figure. The co-ordinates in the notebook are transposed, for instance, env.brown_out = (9,7) (i.e. 10th row, 8th column visually) and should be instead (7,9) to match figure 1.

However this doesnt affect the code, it may be an issue in Task 5 where we need to compare and contrast these 2 grid worlds. Please clarify this discrepancy, thank you!

Hi mizhaan_prajiy_maniy,

Its a effect of how numpy arrays are printed vs how the image is shown in the diagram. Please rotate/transpose as needed for your visualization.

Best Regards,

Ok will make sure, also when you earlier said that the tolerance was 1e-8, did you mean the matrix norm ( || J_{i+1} - J_i|| ) or the max abs ( \max_{s \in S} |J_{i+1}(s) - J_i(s)| )? Thank you!

I think your TAs must have communicated that its supposed to be individual states and not matrix norm.

If you think about it matrix norm isn’t even a valid way, two arrays [0,0,1] and [1,0,0] would have the same norm.

Hi nischith_shadagopan,

We recently made this change, there was no bug, but the small variations in implementations was causing the results to mismatch, so we’ve relaxed the criteria, please resubmit your code.

Due to docker exceptions, can the number of submissions per day be increased to 10?

For Qn2 (Value Iteration), do we need to append J_0(s) in ‘value_grids’ list?

Sir is it possible to increase the submission limit? Since the initial couple of submissions were on wrong tolerance values.

Hi @kbkartik_cs20s020

For submission, you only need the final value grid, the value_grids is provided to help you with the plotting which is subjective evaluation.