Current status of imitation agent in baseline repository

The above script was for doing PPO and IL alternately…

If you want a pure IL, you can try
train.py -ef baselines/custom_imitation_learning_rllib_tree_obs/pure_imitation_tree_obs.yaml --eager --trace

View post 0

Current status of imitation agent in baseline repository

The imitation trainer works. We have generated results for them. You could do a training and simultaneous evaluation using the script
train.py -ief baselines/custom_imitation_learning_rllib_tree_obs/ppo_imitation_tree_obs.yaml --eager --trace
(drop -e flag if you don’t want to do evaluation)
The only thing is the OR expert solution uses was for an older flatland version where the malfunction rate was different. So if you are training with malfunctions, you can workaround it by doing the below changes in the flatland source code

change below line in method malfunction_from_file in the file flatland.envs.malfunction_generators.py

mean_malfunction_rate = 1/oMPD.malfunction_rate

The documentation here https://flatland.aicrowd.com/research/baselines/imitation_learning.html is a bit old , we will update it soon.
You can refer to this Google Colab notebook also which has the details along with the results https://colab.research.google.com/drive/1oK8yaTSVYH4Av_NwmhEC9ZNBS_Wwhi18#scrollTo=P_IMrdL27Ii7
Let me know if you are facing any issues.

View post 2

RLLib Baselines on Colab!

We have taken the repo from https://gitlab.aicrowd.com/flatland/neurips2020-flatland-baselines
and made it into a simple colab notebook

All training scripts are also provided, so one can modify the configs and do runs of their own. Evaluation is also run and the script to calculate scores on an independent test set is also provided.

View post 4

Using a trained agent in RLlib

Your approach seems correct in principle … not sure why the trainer cannot restore from checkpoint. You could compare with the example provided.

View post 0

Using a trained agent in RLlib

you can refer to the rollout.py script in the AIcrowd baselines for flatland

And the corresponding script

Note that this runs small environments with a custom seed. You will have to change the environment logic for your purpose.

View post 0

Local evaluation error: env_client.step called before env_client.env_create() call

Try the steps here
https://flatland.aicrowd.com/getting-started/first-submission.html#troubleshooting

View post 1

Expert demonstrations for Imitation Learning: Recreating Malfunctions

The flatland-rl version has been updated to 2.2.2. (Upgrade it using the command pip install -U flatland-rl). Can you check if the malfunctions are replicable with the same seed? Let us know if you are facing any issues.

View post 1

Expert demonstrations for Imitation Learning: Recreating Malfunctions

It does it slightly different from the MARWIL/Apex-DQfD versions in that it runs every episode alternatively via IL and RL (the ratio is defaulted to 50% ratio but it can be changed and also decayed over time by changing the configs).

View post 0

Expert demonstrations for Imitation Learning: Recreating Malfunctions

You could also try our online RL Solution which does not require any of these intermediate steps like generating experiences. It runs everything on the fly…
You can find a pure IL and IL + PPO solution here

change below line in method malfunction_from_file in the file flatland.envs.malfunction_generators.py

mean_malfunction_rate = 1/oMPD.malfunction_rate

View post 0

Expert demonstrations for Imitation Learning: Recreating Malfunctions

Are you using the same flatland versions for both creation and loading environments? The solution for creating the experiences in the AICrowd baselines for MARWIL and APE-X DQfD were mostly used in environments without malfunctions and they used the seed value of 1001 (https://flatland.aicrowd.com/research/baselines/imitation_learning.html).

View post 0

Solution Codes and Approaches

I haven’t submitted it yet. But I can share the results of few envs from the local evaluation

Evaluation Number : 1
Reward : -43.83333333333334

====================================================================================================
Evaluation Number : 1
Current Env Path : ./test-envs/Test_5/Level_1.pkl
Env Creation Time : 0.42258167266845703
Number of Steps : 1120
Mean/Std of Time taken by Controller : 0.01895513470683779 0.0018288652394336735
Mean/Std of Time per Step : 0.1480530451451029 0.00961317459096298

Evaluation Number : 2
Reward : -43.41666666666668

====================================================================================================
Evaluation Number : 2
Current Env Path : ./test-envs/Test_3/Level_0.pkl
Env Creation Time : 0.29577183723449707
Number of Steps : 960
Mean/Std of Time taken by Controller : 0.02316151708364487 0.002471594250438426
Mean/Std of Time per Step : 0.14834324022134146 0.01293447105111391

Evaluation Number : 3
Reward : -52.00000000000002

====================================================================================================
Evaluation Number : 3
Current Env Path : ./test-envs/Test_6/Level_0.pkl
Env Creation Time : 1.5495717525482178
Number of Steps : 1760
Mean/Std of Time taken by Controller : 0.02436668398705396 0.0039495335690074304
Mean/Std of Time per Step : 0.19943869560956956 0.024274925544436107

View post 0

Solution Codes and Approaches

I have added another code file with a different approach that does not use a model.

The code can be found in the local Github location

For a simple demonstration of how we solve a dense railway network, simply run the file
MultipleAgentNavigationObsConflict.py.
This file does not use any additional packages other than the ones required for flatland and can be run with the latest flatland-rl version 2.1.10

View post 1

Solution Codes and Approaches

I have put up some code here

Adding to Erik’s comments, my observations are

• These models do not show improvement even after training for longer periods and show comparable performance, suggesting that we need to do better feature engineering.

As of now, I next plan to do some visualizations and add documentation to the code to better.

Any comments/suggestions are most welcome.

View post 3

Submission Errors Flatland

Hi @mohanty @ashivani

I am getting an error on evaluation
https://gitlab.aicrowd.com/nilabha/flatland-challenge-starter-kit/issues/6

Can yo help with the logs

View post 0

Submission Errors Flatland

@ashivani

I am not able to see logs but I can see comments on the issue though only the first line.

2019-08-03T15:15:13.396985671Z Traceback (most recent call last):…

View post 0

Submission Errors Flatland

@mohanty

I am getting an error in evaluation
https://gitlab.aicrowd.com/nilabha/flatland-challenge-starter-kit/issues/3

Somehow I cannot see any error logs though debug=True
I have tested this in the local environment (using redis server etc…) and it is working.
Can you please help with the error logs.

Thanks,
Nilabha

View post 0

Submission Errors

Thanks kongas
I used the below code to remove the images

filter_func = lambda x: str(x) not in lsRemove
test_img = (ImageList.from_folder(path).filter_by_func(filter_func))

Though there is another error…
https://gitlab.aicrowd.com/nilabha/snake-species-identification-challenge/issues/33

@mohanty
Has the competition ended or will it restart again. Would have liked to get a score as my validation results were good.
Is it possible to get the error logs?

View post 0

Evaluation stuck [Edit : Evaluation took a long time]

@mohanty

all submissions seems to be queued for a long time.
Is there some problem?

Thanks

View post 0

Submission Errors

@mohanty
I have put a workaround to find the images which fail loading and delete them and later add these probability which are all equal to 1/45
However I still get an error
https://gitlab.aicrowd.com/nilabha/snake-species-identification-challenge/issues/32

Can you please help with the error?

Thanks,
Nilabha

View post 0

Submission Errors

@mohanty @arjun_nemani

Can you please help with the error logs. I keep getting an evaluation failure.
https://gitlab.aicrowd.com/nilabha/snake-species-identification-challenge/issues/31

I have tried reducing the batch size also in case of any memory issue. However the error still seems to be there.
From trial and error I notice that the error happens in the code line learn.tta function from the fast ai library.

Thanks,
Nilabha

View post 0

Submission Errors

@mohanty
I have been using trial and error to see which line of code the submission is failing, unfortunately I have crossed the limit of submission
Could you provide the error logs to debug the issue?
I have isolated the issue to few lines of code but it would be helpful to know what is the error.

Thanks,
Nilabha

View post 0

Submission Errors

@mohanty
I have set num_workers=0
I still get the error.
https://gitlab.aicrowd.com/nilabha/snake-species-identification-challenge/issues/24
Could you share the error logs?

Thanks,
Nilabha

View post 0

Submission Errors

I am feeling guilty having to ask you everytime
I wouldn’t mind if all my logs are made public to be honest.

But on another note would the deadlines of this competition be extended. Due to less time left in the competition, I haven’t done much to replicate this submission process locally (I think there are some instructions in the link https://github.com/stanfordnmbl/neurips2019-learning-to-move-starter-kit). As of now I just do bash run.sh to test that the code is working in my local.

View post 0

Submission Errors

@mohanty
I have redone with a fresh copy from repo and updating the files.
Sorry for asking you again but I still get the error in my latest submission http://gitlab.aicrowd.com/nilabha/snake-species-identification-challenge/issues/22
Do you know what is the error?

Thanks,
Nilabha

View post 0

Submission Errors

@mohanty
run.sh it seems already has executable permission, hence there is no change in my local files on doing chmod +x run.sh
permissions are -rwxrwxrwx in my local
I have looked at the permissions for the other files and they are the same.
Do you think there is some other error?

Thanks,
Nilabha

View post 0

Submission Errors

Thanks. Its working now. The build was also failing due to the environmental.yaml checked using LFS.
Unfortunately the evaluation is failing now after a successful build. Can you kindly help me with the error logs?
https://gitlab.aicrowd.com/nilabha/snake-species-identification-challenge/issues/18

View post 0

Submission Errors

@mohanty

I am trying to submit but I am getting below error
Submission failed : Malformed JSON provided in aicrowd.json
https://gitlab.aicrowd.com/nilabha/snake-species-identification-challenge/issues/15

I don’t see any issues with this file.

Do you know what is the problem?

Thanks,
Nilabha

View post 0

Submission Errors

@mohanty

I have uploaded another version of environment.yml file but the image build fails.

Can you please provide me with the error logs?

https://gitlab.aicrowd.com/nilabha/snake-species-identification-challenge/issues/10

Thanks,
Nilabha

View post 0

Submission Errors

Hi Mohanty

Thanks for the reply. I submitted some new submissions after committing but somehow they are not getting submitted for evaluation.
For example:
https://gitlab.aicrowd.com/nilabha/snake-species-identification-challenge/commits/submission-v3.2

Can you let me know what is the problem?

Thanks,
Nilabha

Thanks,
Nilabha

View post 0

Submission Errors

@mohanty @shivam @arjun_nemani
I can’t quite find any logs available for my failed submission.
https://gitlab.aicrowd.com/nilabha/snake-species-identification-challenge/issues/3

Request you to kindly share the logs

I see many submissions of other participants also failing. In case of an issue in submission do we just message the organisers to get the logs or is there any other way.

View post 0

Cannot upload trained model to gitlab.aicrowd

Its working now after I redid the whole thing and was able to push a submission. However I see the image built successfully but failed after that.I am not able to see any logs. Where can I see the error / log information of why it failed.

[aicrowd-bot @aicrowd-bot added [Failed] label after [image_built_successfully]

View post 0

Cannot upload trained model to gitlab.aicrowd

I am getting an error pack exceeds maximum size

git push aicrowd submission-v0.3
Locking support detected on remote “aicrowd”. Consider enabling it with:
$ git config lfs.https://gitlab.aicrowd.com/*/info/lfs.locksverify true
Uploading LFS objects: 100% (12/12), 146 MB | 3.4 MB/s, done
Counting objects: 71, done.
Delta compression using up to 4 threads.
Compressing objects: 100% (69/69), done.
remote: fatal: pack exceeds maximum allowed size
error: pack-objects died of signal 13MiB | 1.62 MiB/s
error: failed to push some refs to 'git@gitlab.aicrowd.com:.git’

I get same error with git and git-lfs

Any idea what is the issue? Size including weights is less than 200 MB

View post 0

Issue in Test Images

I notice that there are 44 images in the test set of 17331 are not readable. Few example are 1c56622f067257a814254b3a0d66a504.jpg , 2b8060650fbc5c4d04c831e37b6a2af3.jpg
Do we ignore these images? Will these be ignored in the submissions?

Following code gets the list of images:
lsFiles has the full list of paths for the images
from PIL import Image
lsRemove = []
for sFiles in tqdm(lsFiles):
try:
Image.open(sFiles)
except:
lsRemove.append(sFiles)

View post 0

Submission Error Log for Snake Identification Competition

Hi Shivam,

Let me know if you are able to share the logs or where I can find them.
I am keen to try couple more submissions.

Thanks,
Nilabha

View post 0

Submission Error Log for Snake Identification Competition

Hi Shivam

I cant quite find any logs in my gitlab issue.
https://gitlab.aicrowd.com/nilabha/snake-species-identification-challenge/issues/3

I see many submissions of other participants also failing. In case of an issue in submission do we just message you to get the logs or is there any other way.

Thanks,
Nilabha

View post 0

Error/ Evaluation log for Snake Species Competition Submission

This is the link to my submission
https://gitlab.aicrowd.com/nilabha/snake-species-identification-challenge/issues/3

View post 0

Error/ Evaluation log for Snake Species Competition Submission

Hi,

I submitted my model for submission in the snake species competition but it has label as error. Can you send me the error logs or is there any way I can see the error logs?

View post 0

Submission Error Log for Snake Identification Competition

Hi Shivam

I pushed my model to the gitlab rep for submission. However I see the image built successfully but failed after that.
[aicrowd-bot @aicrowd-bot added [Failed] label after [image_built_successfully]

I noticed this thread for the obstacle tower

I am interested in teh error log for
https://gitlab.aicrowd.com/nilabha/snake-species-identification-challenge/issues/3
Do we know how we can get the error logs or you will be able to provide it.

Thanks,
Nilabha

View post 0