Outreachy Round 22 Applicant: Abhipsha Das


I am Abhipsha, an Outreachy applicant from India. I graduated last year with a degree in Computer Science and am interested in ML applications and python development. I am proficient in python and have experience in writing ML code. I noticed PyMC as a potential project before the initial application approval came in, so I am quite happy to be here as I am very interested in this organisation which I didn’t know about before.

Both ‘Improve PyMC3 example notebooks for PyMC3 4.0 and Aesara’ and ‘Integrate Variational Inference with the JAX backend’ are projects that align with my interests. I will probably find out more as I start contributing, so I look forward to working with everyone! :blush:

1 Like

Hi @Abhipsha_Das and welcome!

I am thrilled to see applications start to come in :grinning_face_with_smiling_eyes:

I am the mentor for the “Improve PyMC3 example notebooks for PyMC3 4.0 and Aesara” project, @ferrine is the mentor for “ntegrate Variational Inference with the JAX backend” so I am tagging him as well.

For the pymc-examples project you may have to do some small contributions to pymc3 itself or even Aesara if there were some issues blocking the progress on the notebooks, but work will happen at pymc-examples. I have been working on the contributing guide as well as building up some infrastructure. The contributing guide should have everything you’ll need to find where to work and how to get started, if there is anything not clear enough don’t hesitate to ask here.

It can also be daunting to browse through all the issues and or notebooks, so given your interest in both examples and variational inference, I’d probably recommend going over some of the notebooks on variational inference and updating them where possible in the process. You may have already seen I am behind schedule evaluating the state of each notebook and creating an issue for each with some guidance, but I can prioritize VI notebooks and I have also written some general guidelines on things to look out for in notebooks to see if they are outdated.

1 Like

Thanks a lot @OriolAbril ! It indeed is a little daunting to go through the issues so thank you for the guidance. I will go through them as suggested and see how I can contribute.

1 Like

Let me know if you have any doubts on how to contribute or want some help finding notebooks on specific topics other than VI

1 Like

Hi, I’m a mentor for the VI project, during the contribution process I give a research+development assignment to evaluate the skills. I’ll post it tomorrow, it is about how our VI works)


Sure @OriolAbril I’ll get back to you soon(~1 day) regarding my doubts. Thanks a lot.

Hi @ferrine ! Thanks for the introduction, I look forward to the info on the VI assignment.


There is a Repo GitHub - pymc-devs/pymc-examples: Examples of PyMC3 models, including a library of Jupyter notebooks., which contains notebooks devoted to pymc3. Some of the notebooks are about variational inference (you’ll find them there). You can get a brief idea how should a good notebook look like.

Environment suggestions to easily share the progress with me

  1. Fork the repo with these examples, and create a new branch, e.g. outreachy-name and send me the link by email maxim.v.kochurov@gmail.com
  2. Then choose at one (or more if you have time) f-Divergence from this list f-divergence - Wikipedia (except KL and beta-KL) and create a notebook that describes your f-Divergence and evaluate it on a simple dataset, e.g. 8-schools (centered/non-centered) using an existing Approximation in pymc3 of your choice against KL divergence that is already implemented.
  3. Your implementation should be clean and well explained (or self-explanatory), you can find the way to implement and use the divergence in pymc3 source code
  4. Write me if any questions, I’m happy to help!

Noted, thanks a lot @ferrine !
I’ll get back to you soon with any questions I have! :slight_smile: