NIH | National Cancer Institute | NCI Wiki  

Error rendering macro 'rw-search'

null

Versions Compared

Old Version 7

changes.mady.by.user Unknown User (bearceba)

Saved on

compared with

New Version 8

changes.mady.by.user Unknown User (bearceba)

Saved on

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.
Activities completed this month
  1. I have a ported version of codalab to python3 working with custom changes to python and jQuery to update those libraries...status? -BB
  2. Opened these Challenges:
    1. SPIE AAPM (http://spie-aapm-nci-dair.westus2.cloudapp.azure.com/competitions/4#participate-submit_results) - submissions working, but we need to change open validation phase to open ended phase after test. Will require a code update from main codalab branch....status -BB
  3. Demo docker submissions to Emily Townley who is designing a MIDRC challenge. I should have a non-gpu and a gpu submission ready soon:
    1. I did in fact get a cpu submission working and am fine tuning it. (http://medici-codalab-main.eastus.cloudapp.azure.com/)
    2. I testing briefly with a gpu machine and it works too. 
    3. We are using the forum to go back and forth as people test my tutorial (https://github.com/QTIM-Lab/MedICI-Challenge-Bundles/tree/main/aks_MedNIST_bundle/README). There some interesting bugs that we are working out:
      1. Windows uploaded zip files are not behaving the same as ubuntu or mac based ones.
      2. While not an issue on my system, some parts of the training algorithm require some adjustments and the organizers (MDRC folks) are requesting that I make code changes to the tutorial so that users don't have to figure that out. Sounds reasonable.
  4. With Carolyn's help the Kheops document has been revised to have a more fluid and professional language consistent with reports of this style. Jayashree has commented on the report and her comments are as follows:
    1. what is teh the gap that you are trying to fill
    2. add 1/2 page about what medici is and why you are looking for something like kheops
    3. also the tone of the writing (especially closing thoughts) is pretty conversational - perhaps a more formal tone is preferred for  reports of this sort
    4. i would be careful about “Where we are with challenges, using KHEOPS solves no issue and provides a solution to a problem we don’t really have.”

...