Week 7 - Implementing Inference on federated learning, drafting poster and making presentation slides ahead of ARC Forum meetings

This week, I started to run inference on federated learning to have a visual output of what the results of the model predictions are meant to look like.
Model prediction was not as accurate and distinct as it was expected to. This is because the model took a turn and exploded the general weights of the global model hence leading to it producing just the background image. This is not really ideal for the presentation next Tuesday but I will run it again with more epochs, implement a weight balancing technique and then re run it, hopefully producing better segmentation of eye images
Started making the draft presentation slides and showed them to Miguel, Ruaridh and Zakaria during our group project meetings and 1:1s to receive any feedback prior the presentation.
Recieved some really good feedback and it was encouraged that Mo and I present together as we worked on the same project and used each other’s code hand to hand. It also showed effective collaboration and a proper timeline on how everything went down for the past eight weeks which is really good for presenting

Presenting the project output in the ARC Forum meetings
Clean code, fix merge conflicts with Mo and submit to main branch one last time
Continue making the poster ahead of the celebration event based off result output
Final 1:1s, group meetings, and retrospective meetings
Clear laptop and hand in to David’s desk
Have a mock presentation with Ruaridh, Zakaria and Miguel and recieve more feedback