Question Solved

What are the best practices for writing rebuttal letters during peer review?

Back to Forum
Starting 0 pts 0 followers
National Institute of Technology Karnataka · Posted

Context

I want to start using federated learning for my research on privacy-preserving healthcare analytics. The datasets involve patient records from three hospitals and we cannot share data across sites.

What I am looking for:

  • Python frameworks that are actively maintained and used in research
  • Any frameworks with good support for heterogeneous data distributions (non-IID)
  • Something that has a reasonably gentle learning curve for a researcher (not a systems engineer)

I have looked at PySyft and Flower. Any experience with either, or other alternatives I should consider?

Sign in to join the discussion.

4 Replies

0
Arjun Venkatesan Growing 195 pts · Accepted answer

Great question.I went through something very similar in my second year.

The key insight for me was that LoRA is actually quite well-suited for NER tasks, especially in low-resource settings. I would recommend:

  1. Use LoRA with r=8 or r=16.don't go higher for 8K samples
  2. Apply LoRA to attention layers only, not the feed-forward layers
  3. Use a cosine learning rate schedule with warm-up (10% of steps)

For Telugu-English code-mixed NER specifically, you might also look at MuRIL.it is pretrained on Indian language data and often outperforms XLM-R on Indic tasks even with less fine-tuning data.

0
Replying to Arjun Venkatesan
Xena D'Souza Starting 25 pts · Accepted answer

If you are not already using a reading group format with your students, I would strongly recommend it. Forces everyone to articulate what they read.

0
Replying to Arjun Venkatesan
Shiva Prasad Active 480 pts · Accepted answer

Highly recommend checking out the PMRF portal for the actual numbers.they update the stipend structure annually.

0
Replying to Arjun Venkatesan
Nandita Ghosh Distinguished 720 pts · Accepted answer

Thank you for the honest take. This is the kind of answer I was looking for.not the sanitized version.