iceberg-project / Penguins

MIT License
3 stars 3 forks source link

Feature/rp script #10

Closed AymenFJA closed 5 years ago

AymenFJA commented 5 years ago

This Pull request is adding the Radical Pilot code to the Penguins Repository.

AymenFJA commented 5 years ago

The environment.yaml is to recognize the structure of the code. The EnTK code is not necessary at this moment, we only have one task to be repeated based on the dataset size. So the best approach, for now, is to use RP to submit bag of tasks (BOT).

bspitzbart commented 5 years ago

@AymenFJA Have you executed the rp_script on bridges? Do you have output from test data files?

AymenFJA commented 5 years ago

Dear @bspitzbart , Yes, I did execute the code on bridges, and I have all the output.

bspitzbart commented 5 years ago

Can you send a pointer to the output? Especially the science files.

AymenFJA commented 5 years ago

@bspitzbart Please see below : /pylon5/mc3bggp/aymen/radical.pilot.sandbox/rp.session.js-17-21.jetstream-cloud.org.aymen.018130.0001/pilot.0000/unit.000018/test/v3weakly_unetr_bs96_main_model_ignore_bad/300/GE

bspitzbart commented 5 years ago

Permission denied. Can you make group readable?

AymenFJA commented 5 years ago

I just did, can you check please.

AymenFJA commented 5 years ago

@bspitzbart any updates ? should we approve?

bspitzbart commented 5 years ago

I still could not read the output directory. Can you add read and execute permission throughout the directory tree?

On Aug 26, 2019 8:36 AM, "Aymen Alsaadi" notifications@github.com wrote:

@bspitzbart https://github.com/bspitzbart any updates ? should we approve?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/iceberg-project/Penguins/pull/10?email_source=notifications&email_token=AAN5FWYKKJ2J64MR4AYK3ZLQGPE4RA5CNFSM4IFTMM3KYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD5EHP5Y#issuecomment-524842999, or mute the thread https://github.com/notifications/unsubscribe-auth/AAN5FW7DOYT72PL2TWG7SPDQGPE4RANCNFSM4IFTMM3A .

bspitzbart commented 5 years ago

There are misplaced quotes in the scripts, unit.000000.sh and unit.000001.sh, lines 27-28 export "CUDA_VISIBLE_DEVICES=" export "NODE_LFS_PATH=/local/6220453" Please fix before merging, although they do not seem to affect anything.

AymenFJA commented 5 years ago

Hello @bspitzbart ,

unit.xxxx.sh is the script of the compute unit (task) that is generated by Radical pilot to be submitted. Radical pilot checks the user configuration and set up the script based on that. This will not affect anything.