jyotish

Can I use GPU for prediction?

Hello @woosung_choi @Li_Xue_Han

Sorry, my bad. GPUs are not available for this competition. Apologies again for the confusion.

View post 4

Can I use GPU for prediction?

Hello @Li_Xue_Han

Edit: GPUs are not available for this competition

You can use GPUs for inference. Please refer How to enable GPU for your submission? for more information on using GPUs for your submission.

View post 0

"pod deleted during operation" during generating predictions

Hello @Piteryo

The evaluation ran out of memory. Can you check how much RAM is being used when you run this locally?

Here is a screenshot of the RAM usage for reference.

Screenshot 2021-06-20 at 6.28.09 PM1326×510 65.2 KB

View post 0

Failed at "Build Image" with no logs

Hello @kim_min_seok

Thanks for notifying us about this. This was an issue on our end and it is fixed now. We queued the effected submissions for re-evaluation.

View post 3

Colab Notebook (baseline submission)

Hello @charel_van_hoof

You probably want to post this in https://discourse.aicrowd.com/c/neurips-2021-nethack-challenge/

View post 0

Problems with starter kit setup

Hello @dylan_miller

It looks like one of your apt repositories is giving the error. Kitware also made cmake installable using pip. You can run the following to install the latest version of cmake from kitware.

pip install -U cmake

Reference: https://pypi.org/project/cmake/

View post 1

Discussion on SiamMOT baseline

Hello @octo

Replied on the thread. Let us know if you need any further help on this

View post 1

Getting error in siam-mot baseline submission "pack exceeded maximum allowed size"

Hello @paulids

You need to use Git LFS to push the large files in your repository (your saved models, etc.,). You can find more information on how to use LFS here

View post 0

AssertionError: Incorrect prediction directory structure. View the submission for more details

Hello @SenticLab.UAIC

It looks like # Prediction phase heading is missing in your notebook. Please add this heading after your training code block has ended. We use these to figure out which part of the notebook needs to be executed during evaluation.

For more information please refer this discussion.

View post 1

Dataset: transport endpoint not connected

Hello @e_t

Can you please drop a mail to servicedesk@aridhia.com regarding this?

View post 0

Submission Error : 'DATASET_PATH' not recognised as an internal or external command

Hello @jigsaw

Can you share any output that you might have got after running the command? Do you see something like

[NbConvertApp] Executing notebook with kernel: python

when you run the command?

View post 0

Submission Error : 'DATASET_PATH' not recognised as an internal or external command

Hello @jigsaw

Can you try this?

View post 0

Submission done -- Inference time is too high!

Hello @e_t

The submission command runs your notebook locally once (without the training cells) before submitting it to AIcrowd. This generally should help you find errors in the notebook and avoid wasting your daily submission quota.

You can add --no-verify flag to aicrowd notebook submit command to skip the local evaluation and directly submit your notebook to AIcrowd.

View post 0

Submission Duration

Hello @prachi_nagpal

We run your notebook locally before submitting it to AIcrowd. If the cell is taking too long or fails to run, there is a very good chance for it to fail during evaluation as well.

If you want to submit skipping the local run, you can add --no-verify flag to aicrowd notebook submit command.

If you want to see what exactly is going wrong here,

There should be two notebooks that get executed

1. install.ipynb
2. predict.ipynb.

You can see the name of the notebook that is getting executing from the cell output.

If the cell gets stuck when running install.ipynb, it means that the installation commands you specified at the top of the notebook are causing this issue.

If it’s predict.ipynb that is causing this issue, please re-check the inference code for any blocking calls.

Another useful thing here would be to check if the headings in the notebook are not mistyped. For more information on this, please refer this post.

View post 3

Submission Duration

Hello @prachi_nagpal, @no_name_no_data

Are you facing this issue after making a submission (the notebook runs fine, you get a link to the new submission and you see this in the submission logs) or before it (not able to submit to AIcrowd)?

View post 0

Submission error occured and few logs

Hello @k_k

Shared the traceback over a private message.

View post 0

Submission failed

Hello @ieghor_borisov

Can you do the following and let us know if it worked?

1. Close the existing jupyter lab servers by going to the terminal from which they were started and pressing ctrl + c key combination.
2. Run rm -rf $(jupyter --runtime)/* in your terminal.
3. Start jupyter lab again and try submitting the notebook.

It looks like your jupyter lab’s runtime directory is corrupted. This generally happens when the jupyter lab server doesn’t exit properly (for example force closing the terminal without stopping the server by pressing ctrl+c)

View post 1

Submission failed

Hello @ieghor_borisov

Can you share the outputs of the following commands?

jupyter-lab list
jupyter-notebook list

View post 0

Submission Error(Cloud storage full)

Hello @Rohith-Reddy

Thanks for notifying us about this. We re-evaluated the submission that had this issue.

This was a glitch on our end due to which the submission notebook didn’t load.

View post 1

Error : OSError [107] : Transport endpoint is not connected

Hello @prachi_nagpal

Can you let us know when you are getting this error? Is it after running some aicrowd command? Or when trying to submit?

It would help us understand what went wrong better if you can share a few steps to reproduce this error.

View post 0