Add basic SingleLayerPerceptron Class #220
Description
Right now, the main library lets users train their own custom Torch models by using the Trainer class. However, we want to give users who are not familiar with Torch the possibility to create some basic models.
Taking inspiration from scikit-learn's workflow, we first want to support the creation of SLPs (Single Layer Perceptrons) for both classification and regression tasks. The model should have a simple API (please look at https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.Perceptron.html) and should internally use our Trainers (do not write any training-related code here).
Since this issue will create a new paradigm for Fenn, please ask questions and collaborate with us while working on this. Feel free to join our Discord channel to collaborate with us!
Metadata
Metadata
Assignees
Type
Projects
- Status
Activity
carma439 commentedon Jul 6, 2026
Hello, I am new to fenn. I would like to take this opportunity to work on this issue and get acquainted with fenn.
blkdmr commentedon Jul 6, 2026
Hi @carma439 this issue has been assigned to you! Write your code inside of the nn folder (create there a dedicated model dir). If you want, you can join our discord to collaborate with us!
blkdmr commentedon Jul 13, 2026
Hi @carma439, i have just accepted an MLP class which is very similar to your task. You can use that code as reference
carma439 commentedon Jul 13, 2026
Hey, sorry I got busy at work. I will look at it. Thank you!
samridhibhardwaj006 commentedon Jul 28, 2026
Hi @carma439, just checking in , are you still planning to work on this?
If you've moved on, I'd love to pick it up! @blkdmr happy to use the
MLP class you mentioned as reference if this is available.
carma439 commentedon Jul 28, 2026
Hi @samridhibhardwaj006 please go ahead.
I got really busy at my work and wasn't able to progress much due to time constraints.
Really sorry @blkdmr, I would have to drop off for this issue at the time.
blkdmr commentedon Aug 6, 2026
@samridhibhardwaj006 can i assign this to you?
blkdmr commentedon Aug 6, 2026
@carma439 don't worry! You will have more opportunity to contribute if you want and have time!
NoumanParvez12 commentedon Aug 7, 2026
if no one is working on it . can you assign it to me @blkdmr
blkdmr commentedon Aug 7, 2026
@NoumanParvez12 yours!
blkdmr commentedon Aug 8, 2026
Hi @NoumanParvez12, another contributor opened a PR to solve this problem. However, since you have been assigned to this, I'll not accept his work unless you decide to quit. Take your time, but please let me know if you don't have time to work on this.
NoumanParvez12 commentedon Aug 8, 2026
@blkdmr Yes, I will try my best
NoumanParvez12 commentedon Aug 8, 2026
Hi @blkdmr — quick question before I start: I checked, and sklearn.linear_model.Perceptron only supports classification (no regression variant in sklearn under that name). Since the issue asks for both classification and regression, should I follow the same pattern as MLPClassifier/MLPRegressor and create two separate classes — SingleLayerPerceptronClassifier and SingleLayerPerceptronRegressor — or did you have different naming/API in mind for the regression side?
blkdmr commentedon Aug 10, 2026
You can follow the mlp module structure, and you can use either SLP or Perceptron (classifyifier/regressor), avoid the full SingleLayerPerceptron name
NoumanParvez12 commentedon Sep 1, 2026
Hi @blkdmr — just checking in on this whenever you have a chance to take a look.
Happy to make any changes if needed. Thanks!
ShhlokRastogi commentedon Oct 8, 2026
Hi @blkdmr , I’m interested in working on this issue. I noticed that PRs #309 and #317 were previously opened for it, but the issue is still open. Are you still looking for a contributor to work on this? If so, I’d be happy to take it up.
NoumanParvez12 commentedon Oct 8, 2026
Hi @blkdmr, just a quick update from my side: #317 has been open for a few weeks and is ready for review. All unit tests pass locally, but CI hasn't run because the workflow approval expired. I'm happy to address any feedback or make changes. @ShhlokRastogi , thanks for your interest. Maybe there's another open issue you could take, or you could review #317 if you spot anything.