FedNova

[jafermarq]

FedNova

• Title: Tackling the Objective Inconsistency Problem in Heterogeneous Federated Optimization
• Venue: NeurIPS 2020
• Link to paper: https://proceedings.neurips.cc/paper/2020/hash/564127c03caab942e503ee6f810f54fd-Abstract.html

Do you want to work on this baseline?

🌻 Check everything about the Summer of Reproducibility on flower.dev/summer

All available baselines are listed in the Summer of Reproducibility Dashboard and also in the GitHub Issues with the summer-of-reproducibility label. The content is the same.

📝 It is advised to complete these steps before your start working on your code. But if you can't wait to implement your baseline with Flower (we totally understand it 😄), please ensure you follow the steps on how to contribute a new baseline.

What follows are the steps 1 & 2 in the Summer of Reproducibility instructions.

1. Join the Summer of Reproducibility program

• Join the Flower Slack and say "hi! 👋" in the channel #summer-of-reproducibility.
• Pick a baseline from our curated list <---------------------------------------- [you are doing this now]

2. Define the scope of your contribution

• What are you going to reproduce? Add a comment to your issue and tell us about your plan regarding this baseline: what experiments from the paper are you reproducing?, for which datasets ? the more details you provide us with the better !

• Check if you are eligible for a reward.

  As we have to comply with US/EU regulations, we have checked that individual contributors based on these countries or territories are eligible: Australia, Austria, Belgium, Bulgaria, Canada, Croatia, Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Germany, Gibraltar, Greece, Hong Kong SAR China, Hungary, India, Ireland, Italy, Japan, Latvia, Liechtenstein, Lithuania, Luxembourg, Malta, Mexico, Netherlands, New Zealand, Norway, Poland, Portugal, Romania, Singapore, Slovakia, Slovenia, South Korea, Spain, Sweden, Switzerland, Thailand, United Arab emirates, United Kingdom United States.

  If where you are based is not on the list, please send us an email (summer@flower.dev) letting us know a bit about yourself (where are you currently based?, are you a university student? do you work at a public institution?). Please tell us the baselines you are interested in implementing (i.e. tell us your GitHub issue if you have crated one). We will reach back to you.

• We will discuss with you about your contribution plan, if it sounds like a substantial enough contribution according to the Summer of Reproducibility rules (check our website flower.dev/summer), you'll get the OK from us to start working on your baseline!

What happens next?

• This item will be moved to the In Progress stage by a member of the Flower Team.

• Follow the instructions for creating a new baseline which will guide you through the process step-by-step.

Is something wrong or not clear ?

• Ask a question directly in your issue.
• Reach out to us via the Flower Slack and ask your question in the #summer-of-reproducibility channel
• Check all the details (including FAQ) in the Summer of Reproducibility website: flower.dev/summer

[ashdtu]

Hi @jafermarq I would like to implement the FedNova baseline in the Flower framework.

Contribution plan

• The main contribution will be to implement the Normalized Averaging technique proposed in the paper. The standard techniques of FedAvg and FedProx can be construed as a special cases of this method by setting specific values of certain parameters(such as alpha). Hence implementation of this method would allow the community to compare performance of multiple averaging techniques from a single endpoint in case of heterogeneous datasets.

• Further contribution will be to add support for multiple local client solvers(vanilla SGD, proximal SGD, SGD with momentum)

Experimentation/Datasets
I plan to reproduce experiments proposed in the paper on the Synthetic(1,1) dataset (with alpha, beta = 1) and further on a Non-IID version of CIFAR-10 dataset. The experiments would involve performance comparison between FedAvg, FedProx and this work FedNova).

My Background
I am currently a MSc Computer Science student at MILA-Quebec AI Institute, Montreal. It is one of the leading institutes in AI research. I am based in Canada and hence in the list of eligible countries.

Would love to discuss further and get started. I am on Flower slack as(@ ash, Aasheesh Singh)

[jafermarq]

Hi @ashdtu, it's great to hear you want to reproduce the results in FedNova. Also, thanks for the very complete message addressing all the points. Just one small clarification about your contribution plan is to replicate Figure 5 and Table 1 ?

[ashdtu]

Yes @jafermarq that's correct. Fig 5 corresponds to Synthetic(1,1) experiment and Table 1 for CIFAR 10 dataset.

[jafermarq]

Great! thanks for the clarification. I've ✅ all the points above for Steps 1 & 2 and add you as the assignee to this issue. I also moved this FedNova baseline to "In Progress". You can find first steps on how to start working on your baseline via the link in the What happens Next section in the issue description.

Looking forward to see FedNova in action!

[ashdtu]

Draft PR : #2179 for progress

[jafermarq]

Congratulations @ashdtu your baseline has been merged. We'll follow up with you soon to complete your Summer of Reproducibility journey 😄 !