16 Jan 2025
Day 30: Efficient Pruning for Machine Learning under HE
-Pruning boosts traditional ML efficiency but not in HE-based ML. Tile sparsity enables efficient efficient pruning by skipping entire blocks of encrypted data, reducing operations.
fhe.org/meetups/041-Efficien…
#30DaysOfFLCode
1
2
120
14 Jan 2025
Day 29: Concrete ML - Machine Learning on Encrypted Data
A model is converted to an FHE-compatible format via quantization, and its weights are encrypted. It then performs predictions on FHE-encrypted data, with the results being encrypted
youtube.com/watch?v=DP_4OBNf…
#30DaysOfFLCode
2
142
11 Jan 2025
Day 26: Private Information Retrieval
PIR allows a user to send an encrypted query to a server, which then performs homomorphic computations to retrieve the desired information, sending back the result still encrypted to preserve privacy.
machinelearning.apple.com/re…
#30DaysOfFLCode
2
68
9 Jan 2025
Day 25: Problems and Research Trends in Federated Learning
I read about problems in FL which requires further research. This includes improving efficiency and effectiveness, reduction in training biases, and robustness against attacks. drive.google.com/file/d/1QGY…
#30DaysOfFLCode
2
64
8 Jan 2025
Day 24: Cross-Silo Federated Learning
-It involves fewer clients (organizations) with specific identities
-Most clients participate in each FL round with very few going offline.
-Data Partition can be vertical or horizontal across clients. drive.google.com/file/d/1QGY…
#30DaysOfFLCode
2
86
7 Jan 2025
Day 22: Cross-Device Federated Learning at Apple
Apple uses differentially private federated learning to personalize Siri. This enables an iPhone to respond exclusively to its owner’s voice, even in the presence of other phones and voices technologyreview.com/2019/12…
#30DaysOfFLCode
3
77
5 Jan 2025
Day 21: Privacy Preserving AI - Lecture (Andrew Trask - OpenMined)
Focuses on how privacy enhancing technologies such as differential privacy, multiparty computation and federated learning enables remote training of models on private data.youtu.be/4zrU54VIK6k?feature…
#30DaysOfFLCode
3
93
5 Jan 2025
Completed the #30DaysOfFLCode Learned so much and advanced mixing of PHEs project. Next up: zk proofs for verifiable computations to enhance PETs. Grateful for the journey and the amazing people I met! 🚀 the journey will continue …
1
4
180
4 Jan 2025
Day 20: Beyond Privacy Trade-offs with Structured Transparency
Participants can't prevent other collaborators from misusing their data (copy problem). The proposed techniques enable collaborators to reduce the risks of collaboration. arxiv.org/abs/2012.08347
#30DaysOfFLCode
4
58
2 Jan 2025
Day 18: Federated Fine-Tuning of LLMs (Part 2)
Today, I began with centralized fine-tuning of EleutherAI's Pythia 70M model using the Med Alpaca dataset, which is formatted in a Question-and-Answer style. learn.deeplearning.ai/course…
#30DaysOfFLCode
3
128
30 Dec 2024
Day 15: Secure Multi-Party Computation Use Cases
2. Privacy-preserving ML and Data Analytics
MPC can be used to run ML models on data without revealing the model parameters to the data owner, and vice versa.
eprint.iacr.org/2020/300.pdf
#30DaysOfFLCode
4
62
28 Dec 2024
Day 13: Federated Learning and Privacy
I studied the essence of adopting data minimization and data anonymization principles to ensure privacy in federated learning and federated analytics.
dl.acm.org/doi/pdf/10.1145/3…
#30DaysOfFLCode
4
50
22 Dec 2024
Day 30: Focused on parameter selection and fine-tuning randomizer values for seamless transitions between ElGamal (mod p) and Paillier (mod n) in a mixed PHE setup. Ensuring security, compatibility and efficiency in operations is crucial for mixing PHEs #30DaysOfFLCode
1
4
123
21 Dec 2024
Day 29: Progressed on my MixPHE project:
Refactored ElGamal & Paillier key-handling with OOP.
Modified aggregator class for mix of all homomorphic operations.
Verified encryption, decryption, and transitions between schemes.
Excited to integrate it to syftbox! #30DaysOfFLCode
5
102
20 Dec 2024
Day 30 of #30DaysOfFLCode: I made sense of the FedCKE algo, and recapped the last 30 days journey. Of course, I learned a lot, but more importantly, I made some good friends along the way. Thanks @openminedorg, @iamtrask for this challenge.
GitHub: github.com/Spartan-119/30Day…
1
4
71
Day 30 of #30DaysOfFLCode
Spent the last 24 hours testing the FedRAG project and trying to integrate a feature to the frontend. While I couldn’t finish everything I planned (thanks to laptop overheating and sync issues 😅)
(1/n)
2
3
185
20 Dec 2024
#30DaysOfFLCode Day 30: Last day!
I attended the 3rd Show & Tell by @openminedorg full of fun presentations by fellow participants.
I'm notorious for not keeping a habit for long, so I'm super glad to have finished something successfully for once😋
An excellent learning exp 🔥
1
6
109
20 Dec 2024
Day 28: Attended the last show and tell of #30DaysOfFLCode by @openminedorg . Briefly presented about our Syftbox app using Mix of PHEs
4
77
For Day-30 of #30DaysOfFLCode ,
I attended the final show and tell session organised by @openminedorg . The members of the FedRAG project team demo-ed our syftbox api to the community (alongside some really cool demos by other community members!)
2
3
2
367
20 Dec 2024
Day 30 of #30DaysOfFLCode!🎉🎊
Challenge completed! A huge thank you to the @openminedorg community for the incredible resources and support. These 30 days have been productive and rewarding.
This is just the beginning of a long path toward mastering privacy-preserving AI!
1
9
93