doudi25

Algerian-nlp-dataset-classification

BackProp-Ninja-with-binary-classification-task

i impelemented backprop manually for the BCE loss , to understand in depth how the gradient is calculated

Chest-X-Ray-classification-

new dataset for medical application consisting of different type from medimeta , you can use the mnist type or the original type, applyinig multiclass classification using pytorch and oop model

DeepAutoEncoders-

applying-vanilla-and-convolutional-autoencoder-to-Mnist-dataset

DeepImageGenerator

using Convolutional Autoencoder to generate images

diffusion-tutorial-

i built simple diffusion tutorial that shows how to use miniUnet to denoise images and train the model to do it .

Discriminant-analysis

i implemented gaussian discriminant analysis from scratch with numpy

gan

Generative-Adversarial-Networks

Gans implementation with pytorch , the model was trained on cifar dataset .

Sentiment_analysis

We employed a Long Short-Term Memory (LSTM) neural network to analyze sentiments in a Twitter dataset. The objective was to categorize tweets as either positive or negative based on their content.

doudi25/DeepImageGenerator

using Convolutional Autoencoder to generate images

doudi25/Sentiment_analysis

We employed a Long Short-Term Memory (LSTM) neural network to analyze sentiments in a Twitter dataset. The objective was to categorize tweets as either positive or negative based on their content.

doudi25/Algerian-nlp-dataset-classification

doudi25/BackProp-Ninja-with-binary-classification-task

i impelemented backprop manually for the BCE loss , to understand in depth how the gradient is calculated

doudi25/Chest-X-Ray-classification-

new dataset for medical application consisting of different type from medimeta , you can use the mnist type or the original type, applyinig multiclass classification using pytorch and oop model

doudi25/DeepAutoEncoders-

applying-vanilla-and-convolutional-autoencoder-to-Mnist-dataset

doudi25/diffusion-tutorial-

i built simple diffusion tutorial that shows how to use miniUnet to denoise images and train the model to do it .

doudi25/Discriminant-analysis

i implemented gaussian discriminant analysis from scratch with numpy

doudi25/gan

doudi25/Generative-Adversarial-Networks

Gans implementation with pytorch , the model was trained on cifar dataset .

doudi25/generative-pretrained-transformer

i have trained Nano gpt on Les Misérables Novel by Victor Hugo

doudi25/Geometric_Deep_learning

Using geometric_deeplearning on wikipedia_netowrk with different architectures : Graph_Convolutional_Network and Graph_Attention_Network

doudi25/GPT2-from-Scratch

i implemented gpt2 from scratch end to end and load the weights from hugging face and implement decoding strategy to play with the model

doudi25/Graph-machine-learning

Node classification using Node2Vec method for karate_club dataset

doudi25/KNN-from-scratch

i have implemented KNN classifier from scratch using numpy

doudi25/Mini-Llama2

custom impelementation of llama2 without kv cache .

doudi25/Mnist-Classification-with-pytorch

i use the Mnist dataset for classify handwritten digits with the Pytorch framework

doudi25/Neural_Network-with-Numpy

i built a neural network from scratch with 3 hidden layers amd i have implemented the backpropagation manually

doudi25/PaliGemma-VisionLanguageModel-

PaliGemma 3B model , implemantation with pytorch

doudi25/Quantization

in this repo , you will find the quantization concept in details , including symmetric , asymmetric quantization , quantization aware training , post training quantization

doudi25/Scientific-machine-learning-PINN-

in this project , i have implemented neural network that can solve ordinary differential equation without using any data point just the laws of physics with the pytorch framework.

doudi25/signal-processing-compressive-sensing

applying compressive sensing algorithm to reconstruct signal from random selections .

doudi25/Stream_App

doudi25/SwiGlu-manual-implementation

i implemented the SwiGLu used in the feedforward method of Llama2 , and i apply its gradient manually

doudi25/Variational-Autoencoder

VAE implementation from scratch with pytorch , with tiny imagenet dataset , you can change the dataset but be carefully about the conv output dims , you must change them to fit with your dataset