NovaMind - Stardance

@e returned your project · about 1 month ago

This needs changes before we can consider it shipped. Review the Shipwright feedback, make updates, then ship again.

Feedback

mrrp mrrp~ :3
i'm rejecting this project for the following reasons:
- your ai declaration is entirely ai generated, please write it yourself
- your readme is almost entirely ai generated, please rewrite it yourself
- your demo link does not work, please fix it
we cannot accept your project until these issues have been addressed and resolved. if you have any questions, please open a ticket in #ask-the-shipwrights

Shipwright walkthrough

@kaboom returned your project · about 1 month ago

This needs changes before we can consider it shipped. Review the Shipwright feedback, make updates, then ship again.

Feedback

Hello,
Please address the previous reviews' feedback.
If you have any questions, DM @kaboom or create a ticket in #ask-the-shipwrights.

Shipwright walkthrough

@kaboom returned your project · about 1 month ago

This needs changes before we can consider it shipped. Review the Shipwright feedback, make updates, then ship again.

Feedback

Hello,
I appreciate you rewriting your README by hand, however you've not re-written any of the code, it's still the AI-generated code from earlier.
If you have any questions, DM @kaboom or create a ticket in #ask-the-shipwrights.

Shipwright walkthrough

@Floppy returned your project · about 1 month ago

This needs changes before we can consider it shipped. Review the Shipwright feedback, make updates, then ship again.

Feedback

Cool project! but the canvas seems to glitch out when I clear it (see the vid) please fix it and resubmit!

Shipwright walkthrough

@Rohan returned your project · about 1 month ago

This needs changes before we can consider it shipped. Review the Shipwright feedback, make updates, then ship again.

Feedback

Hey there!
Cool project however It seem's your Ai declaration is AI generated. can you please handwrite it and reship.
Dm @Rohan for any questions : D

Shipwright walkthrough

@kaboom returned your project · about 1 month ago

This needs changes before we can consider it shipped. Review the Shipwright feedback, make updates, then ship again.

Feedback

Hello,
You appear to have rewritten the README with more AI, please do not do that, rewrite by hand.
Also, removing the AI comments and docstrings without actually re-writing code with not be accepted.
If you have any questions, DM @kaboom or create a ticket in #ask-the-shipwrights

Shipwright walkthrough

@kaboom returned your project · about 1 month ago

This needs changes before we can consider it shipped. Review the Shipwright feedback, make updates, then ship again.

Feedback

Hello,
Please address the previous reviews' feedback.
If you have any questions, DM @kaboom or create a ticket in #ask-the-shipwrights.

Shipwright walkthrough

@kaboom returned your project · about 2 months ago

This needs changes before we can consider it shipped. Review the Shipwright feedback, make updates, then ship again.

Feedback

Hello,
Please address the previous review's feedback, we didn't reject based on the docs/ folder.
If you have any questions, DM @kaboom or create a ticket in #ask-the-shipwrights.

Shipwright walkthrough

@kaboom returned your project · about 2 months ago

This needs changes before we can consider it shipped. Review the Shipwright feedback, make updates, then ship again.

Feedback

Hello,
Your project codebase is almost entirely AI-generated with no human effort whatsoever, please re-write majority of your codebase including your README by hand for approval.
If you have any questions, DM @kaboom or create a ticket in #ask-the-shipwrights.

Shipwright walkthrough

Ship #1 Changes requested

@omsatyam987 on NovaMind · about 2 months ago

Built a neural network from scratch using only NumPy — no PyTorch, no TensorFlow, no autograd. Trained on MNIST (60k images), 97.91% test accuracy.

The challenging part was getting backpropagation right. Specifically, the Softmax + CrossEntropy backward pass — the combined gradient simplifies to (predictions - y_true) / batch_size, which means CrossEntropy needs no backward method at all. That only becomes obvious when you’re writing the math yourself instead of calling .backward().

Most proud of the fact that every single component — forward pass, backprop, weight updates, L2 regularization, Xavier initialization, the MNIST binary parser — is written by hand. Nothing is hidden behind a framework.

To test: draw any digit (0–9) on the canvas and click Predict Digit. Draw it centered and reasonably large for best results. The model runs entirely in the browser using weights exported from the trained NumPy model.

1 devlog
10h

Try project → See source code →

Open comments for this post

@omsatyam987 on NovaMind · about 2 months ago

10h 0m 35s logged

Built NovaMind — a multi-layer perceptron trained on MNIST, written entirely from scratch using only NumPy. No PyTorch, no TensorFlow.
What I implemented by hand:

Forward and backward pass through every layer
ReLU and Softmax activations
Cross-entropy loss
Mini-batch SGD with L2 regularization (λ=0.0001)
Xavier weight initialization
Full MNIST data pipeline — download, parse IDX binary format, normalize, one-hot encode

Final results: 99.69% train accuracy, 97.91% test accuracy.
Also built:

Visualization suite — loss curve, confusion matrix, and a prediction grid showing wrong vs correct predictions
Live interactive demo — draw a digit on a browser canvas, the model predicts it in real time using trained weights exported as JSON and a forward pass written in JavaScript

GitHub: omsingh-19/NovaMind
Live demo: NovaMind MNIST Digit Recognizer