Skip to content

Feature Addition (regarding Modularity / Running Multiple Models Simultaneously / Aggregator Direct-Inference / Docker containerization / Pre-pulling policies)#84

Merged
royshadmon merged 127 commits into
mainfrom
pre-main
May 27, 2025

Conversation

@ejbrannonwu
Copy link
Copy Markdown
Contributor

@ejbrannonwu ejbrannonwu commented May 7, 2025

  • Modularity for Aggregator/Nodes (commands require extra parameters; init-command now requires index, module, module_file & db_name, all subsequent commands just require index)
    • module must be the name of the class in the module_path file (ex. MnistDataHandler)
    • module_file is appended to the directory path TRAINING_APPLICATION_DIR that is mentioned in the .env's, edgefl/platform_components/data_handlers
      (ex. mnist_data_handler.py)
    • db_name must be what postgres contains, data in the DB must relate to the module (ex. mnist_fl)
  • Tested running multiple models simultaneously (same data-handler)
  • Tested running multiple models simultaneously (different data-handlers - winniio/mnist)
  • Enabled the aggregator to run direct-inference (new endpoint command, example/explanation shown in README)
  • Pre-pulling policies in the policy listener function for aggregator
  • Docker Containerization
  • Minor fixes:
    • Updated README.md with command updates
    • edited logger messages
    • checking for duplicate index
    • fixed sub-directories
    • moved db checking to init
    • connection to specified db_name in init if edgelake node doesn't have a connection to that DB - connect_to_db() (needs to have had data been connected in postgres already)
    • fixed default .env files to work with module_file
    • Add a message at the end of a training process (via logger) to indicate the training has finished
    • check that module_path exists (if not, don't initialize)
    • node checked to see if actually initialized, clarifying message if not (used to show success even if off)

DDublue and others added 30 commits February 10, 2025 16:53
…ency (manually for now requires changing ModelUpdate to LocalModelUpdate in lib/python3.9/site-packages/ibmfl/aggregator/fusion/iter_avg_fusion_handler.py)
…of ModelUpdate, removing its IBM dependency (manually for now requires changing ModelUpdate to LocalModelUpdate in lib/python3.9/site-packages/ibmfl/aggregator/fusion/iter_avg_fusion_handler.py)
IBM Removal of ModelUpdate/DataHandler
Created a base aggregation model class + federated average agg. model…
Removed IBM federated learning & upgraded platform to Python 3.11
Revert "Removed IBM federated learning & upgraded platform to Python 3.11"
Upgraded platform to run on Python 3.12 + requirements; updated README
…ls-branch

Revert "Revert "Removed IBM federated learning & upgraded platform to…
Merging Logger Branch to Pre-main
MiguelMasc and others added 6 commits May 9, 2025 13:38
…h existed; checked if node actually initialized; started aggregator direct inference; some code cleaning
…list of labels/predictions); enabled aggregator to have an fl_model (in both data handlers); updated direct_inference in the mnist data handler (will update for winniio later)
@ejbrannonwu
Copy link
Copy Markdown
Contributor Author

edit 05/15:

  • added additional feature regarding aggregator running direct-inference and some more minor fixes (bottom 3 above)

@ejbrannonwu ejbrannonwu changed the title Feature Addition (regarding Modularity / Running Multiple Models Simultaneously) Feature Addition (regarding Modularity / Running Multiple Models Simultaneously / Aggregator Direct-Inference) May 15, 2025
@ejbrannonwu ejbrannonwu changed the title Feature Addition (regarding Modularity / Running Multiple Models Simultaneously / Aggregator Direct-Inference) Feature Addition (regarding Modularity / Running Multiple Models Simultaneously / Aggregator Direct-Inference / Docker containerization / Pre-pulling policies) May 27, 2025
@ejbrannonwu
Copy link
Copy Markdown
Contributor Author

edit 05/26: added to pre-main

  • Pre-pulling policies in the policy listener function for aggregator
  • Docker Containerization

Copy link
Copy Markdown
Owner

@royshadmon royshadmon left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

tested and lgtm

@royshadmon royshadmon merged commit 066f1dc into main May 27, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

5 participants