-
Notifications
You must be signed in to change notification settings - Fork 7
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[DOC] Adds Iñigo's blogposts of Weeks 1 and 2 #39
Conversation
🪓 PR closed, deleted preview at https://github.com/dipy/preview-html/tree/main/dipy.org/pull/39/ |
@skoudoro I cannot request review from other participants, am I overlooking something or is there anything wrong with my permissions? |
done, not sure why you can not, I might look at it this weekend |
Why I see kaustav blog posts? can you rebase or merge master to see only your changes. Thank you |
You blog post failed to compile. see here: https://github.com/dipy/dipy.org/actions/runs/9415155137/job/25935562148?pr=39#step:7:216
Please, can you update? That's why we have no preview ps: we also need to update the CI. it should come back as an error |
6755aed
to
b2d1301
Compare
Done, I dropped his commits and squashed mines. Thanks |
Updated. Should be fine now. I do not get the CI notifications, IDK why.
Perhaps I will look into this next week. I opened an issue #40 |
b2d1301
to
81d7cee
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Not sure if this is the latest commit preview:
https://dipy.github.io/preview-html/dipy.org/pull/39/posts/2024/2024_06_07_Inigo_week_1.html
But two of the links are not rendered correctly.
Even for the first one, if I'm not wrong, a single underscore is usually enough.
Also, this blog post is mixing work done during week 1 and week 2, so please split those. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Hi, nice work you've done :)
I found some small RST syntax issues but that's all. Approving ✅
|
||
What I did this week | ||
~~~~~~~~~~~~~~~~~~~~ | ||
On my last post, I wrote about translating the AutoEncoder architecture from PyTorch to TensorFlow. I have successfully done so, and I have attempted to overfit the model with data from the `FiberCup dataset <https://tractometer.org/fibercup/home/>`__ to check whether the architecture is working as intended, and that the training objective is correctly defined (i.e. the loss function is correctly implemented). |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Fixed, thanks
|
||
What is coming up next week | ||
~~~~~~~~~~~~~~~~~~~~~~~~~~~ | ||
With the help of my mentors, we identified possible improvements to the AutoEncoder training process. Yesterday I investigated how PyTorch weights are initialized in convolutional kernels and in Keras Dense layers using the `He Initialization <https://paperswithcode.com/paper/delving-deep-into-rectifiers-surpassing-human>__`. Using custom initializers, one can mimic the same behavior in TensorFlow, which I started to implement also yesterday. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Fixed, thanks
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Dates and contents still mix week 1 and week 2. Can this be addressed, please?
First Week into GSoC 2024: Building the AutoEncoder, preliminary results | ||
======================================================================== | ||
|
||
.. post:: June 7 2024 |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
There is still confusion in dates and weeks. If I'm not mistaken there was a blog post (the first one) that corresponded to the community bonding period (week prior to May 27), and since then there have been two weeks: week 1 (ended Fri May 31), and week 2 (ended Fri Jun 7). So I find that this blog post still confuses dates and contents. Can this please be addressed? 2 blog posts are requested. To avoid this in the future, best is to allocate 30-45 minutes every Friday to write about the work that week. I know it is the most scientifically interesting part for the contributor, but it is important for the mentors, to keep track of the project, and the program.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
You are right, I messed up the dates. I will split the info into two blogs. Sorry for the confusion.
:category: gsoc | ||
|
||
|
||
What I did this week |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Hi @itellaetxe!
Nice blog, clear and concise, Great job.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks Wachiou!
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
A minor comment inline.
What I did this week | ||
~~~~~~~~~~~~~~~~~~~~ | ||
I finished becoming familiar with the TensorFlow + Keras basics and I wrote the training loop and a couple of scripts for instantiating and training the AutoEncoder. | ||
Data loading was also addressed and I am able to load the data from the FiberCup dataset in .trk format using `NiBabel <https://nipy.org/nibabel/>`__, transform it into NumPy arrays, and feed it into the network. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Again, a single underscore is enough in links: https://www.sphinx-doc.org/en/master/usage/restructuredtext/basics.html#hyperlinks
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
fixed, thanks :)
pinging @deka27 |
Two weeks progress. Looks fine to me. sorry for the delayed check. If squashing isn't required, GTG. |
* [DOC] Adds blogpost Iñigo's blogpost of Week 1 * FIX: RCorrects bad RST URL syntax & deletes mentions to week 0 work * DOC: Adds Iñigo's week 1 and 2 blogposts * FIX: Single underscore instead of two for RST URLs 6219f3c
Adds @itellaetxe blogposts for Weeks 1 and .