SOURCE CODE: Mid-Level API

Part 1 (2020) Study Groups
29

Thanks! That sounds like a great suggestion. Are you available to join the zoom call? :slight_smile:

30

Sure, give me one moment

31

:smile: I’m just finishing FastaiV3-part2 and the Xresnet implementation blew me away.
Maybe we can greenlight this idea for this blog post series and co-write it?
Obviously you know the library better than me but we can discuss of a global structure and visuals to produce and split the tasks?

32

I will join in a few minutes. I’m in another fastai study group :sweat_smile:
Thanks for organising @arora_aman

2 Likes
33

Excellent, we’ll wait for you.

34

Thank you, Aman.

Everyone, also please keep an eye out for the SG by Jeremy here: https://forums.fast.ai/t/official-project-group-thread/65817

1 Like
35

That’s fine with me (6pm in France)

1 Like
36

Excellent !

37

Interested

38

Concerning the organisation of the work, I might have a proposition:

Maybe we could assign each other a specific part of the course’s source code to explore and then each day:

  • Have one of us present and explain what he could understand from the source code (sharing his screen and going through the code and maybe sketches to help understand some mechanics (much like Jeremy does))
  • Have everyone ask questions/comment/correct
  • Discuss of a blogpost subject to illustrate the code discussed or even a mini-project + blogpost.

Much like a book reading club but even more uncool !

39

Sorry, is the zoom meeting closed?
I’m not able to join it.

1 Like
40

Sorry Sanyam, I just closed the meeting. Give me two minutes and I’ll post an update here on what was discussed and decided.

2 Likes
41

As a group, we discussed on how to approach towards digging into the source code. We discussed the common problems we face, and decided that looking into the mid level APIs is the best approach to start with as it is highly flexible and can be also applied to other applications such as text/collaborative filtering/segmentation

We want to take a linear approach and gradually build on top each week by digging more into the API slowly.

We will start by looking into the PETs dataset first and gradually start looking into other perhaps more interesting datasets such as imagewang, imagewoof etc

We will split resources and decide at the end of each lecture on who works on what based out of interest, then go away for 2-3 days, complete the individual pieces of work, write multiple small blog posts on the different things we all worked on and come back in 3 days time again as a group to share the information and present to the group our individual findings. This way we all learn and progress together as a group and are able to cover more ground than we would have been able to if we would have worked individually.

1 Like
42

@jeremy could you please make this post a WIKI when you get a chance :slight_smile:

43

Petition to rename the group “Mid-Level API Study Group”? Or “SOURCE CODE: Mid-Level API”

(Or DataBlock API even)

1 Like
44

Granted :wink:

45

I’d like this but I seem to have run out of my like balance… that’s a first :slight_smile:

1 Like
46

Sound good to me.
Maybe we could meet on the Wednesdays to wrap-up the lesson and discuss the assignments.
And meet again on Fridays or Mondays, before the next lesson, to present our work?

1 Like
47

Haha, I wasn’t aware there’s a like balance on the forums. That’s new.

48

Yes thank you for the suggestion. That’s the idea but we haven’t decided the specifics yet. We will be organising another zoom call very soon to discuss the exact timings based on what works for everyone.

I will post the details for the next zoom call soon :slight_smile: