Study Group in Bangalore (Slack & Meetups)

Hi everyone.

We’ll try to record the sessions so we can make it available online later.

Hi @aakashns, We started an deep learning study group for fastai a month back. I thought you guys were doing an online meetup. Also, we will be completing part 1 v2 this week. And discussed to take up part 2 v2 in the morning sessions and part 1 v3 in the afternoon sessions.

Sorry for the late notice.
If you are looking for a offline meetup, let’s collaborate, it would be beneficial for the whole community.

2 Likes

Happy to collaborate! :slight_smile:

We’ve already planned a meetup for this Saturday, so we’ll probably go ahead with that. Do keep us updated with your plans for Part 1 V3

3 Likes

I would like to join this study group and meet all in person…

That’s great Saurabh! Do join for the meetup on Sat, Oct 20

Will be in Bangalore this weekend, look forward to meet all :slight_smile:

hi all,
i am from Bangalore and looking forward to join a study group

1 Like

Hi @aakashns… Isn’t the meetup full for this Saturday in Bangalore? If not please let me know.

1 Like

Seating for the talks is full, but you can attend the mixers at 4:30 pm, to meet other folks.

1 Like

Guys, I have been using Tensorflow to build my project. But for this course, they are using Pytorch. So I started to learn Pytorch lately. Wrote this tutorial to use Pytorch efficiently to build model using torch.nn classes and functions. So check out this brief tutorial about Pytorch’s torch.nn module.

1 Like

Nice! You should consider turning it into a Kaggle kernel, so people can run the code as they read. I wrote a similar kernel recently.

Thanks for checking it out. Yeah great idea I will creat a kaggle kernel for it.

m too from blr… looking forward for rich discussion and learning from the group

Hey guys,
Here’s a live stream of the meetup: https://youtu.be/FWZ6s_-jgDw

1 Like

FastAI Live Study Group Meetup

Sat, Oct 20, 2018

26 people attended the meetup today. We began with a round of introductions, followed by excellent technical sessions by @crazydiv and @vinodreddyg . This was followed by an open discussion on several other topics: FastAI in production, FastAI vs. Keras/Tensorflow, Python vs. R/Matlab, how to approach the course, how to find ideas for projects and much more.

Session Videos:

  1. Productivity Tips for Kaggle Competitions (Kartik)
  2. Gradient Descent & Optimization Algorithms (Vinod)

Plans for future meetups:

  1. Show and tell (blog posts & projects from members)
  2. Discussion of FastAI lecture (doubts, ideas etc.)
  3. Discuss Kaggle competitions and form teams
  4. Technical sessions & paper discussions

Some ideas for future sessions:

  1. Gaining insights from Exploratory Data Analysis (EDA)
  2. Structured way of finding the best hyperparameters
  3. Discussion of Leslie Smith’s paper on Learning Rates
  4. Deploying PyTorch and FastAI models to production

Shout-out to Saurabh Vyas for finding the venue (CoWrks, RMZ Ecoworld).

Also, thanks @jeremy for creating the FastAI course and library, and making it accessible to people around the world! We’re really excited to be a part of the course. :slight_smile:

6 Likes

Looks great! Can you tweet, and share the handle.

Sure! https://twitter.com/aakashns/status/1053711469710204928

Thanks for sharing! You might want to also reach out to the local WiMLDS meetup to get some of them involved too. :slight_smile:

10 Likes

That’s a great idea. Didn’t know about WiMLDS. Will definitely get them involved in future meetups.

2 Likes