Official PyTorch implementation of Extract Free Dense Misalignment from CLIP (AAAI'25)

🧠 A flexible machine & deep learning framework built from scratch using only NumPy

Developed a Marathi speech-to-text application using the Hugging Face whisper ASR models. Trained the model with a custom audio dataset and fine-tuned it for optimized performance. Deployed the model on the Hugging Face Model Hub, achieving a WER of 0.74 for the base model.

predicting Protein Subcellular Localization from quantitative label-free imaging with phase and polarization

🏞 A content-based image retrieval (CBIR) system

Adaptive generalized review summarization system using optimized transformers.

Ordantis @ EXIST 2026 Task 2 — GEMF: Gemini-enriched multimodal fusion for sexism characterization in memes (CLEF 2026).

Embedding analysis and some insights on the GPT-2 architecture

FeelBack es un proyecto personal que consite en una aplicación web simple tipo chat construida con Flask y un modelo de HuggingFace que permite a los usuarios analizar el sentimiento de textos en tiempo real.

Analyse automatique des réponses à des formulaires numériques. Transforme des questions et réponses ouvertes en insights exploitables grâce à l’IA. Inclut détection d’objectif, résumé automatique et chatbot interactif.

Detect sexism in memes using Gemini-Enriched Multimodal Fusion. Rank-leading system for the EXIST 2026 task on sexism characterization.

💳 Detect fraudulent payment transactions with this transformer-based system, leveraging LoRA for efficient fine-tuning and providing real-time predictions.

Analyze the emotional tone of any text using AI and Transformers.

mengkelompokan gambar bedasarkan kemiripan pada sebuah gambar

NLP pipeline using Hugging Face transformers, fine-tuned for text tasks. Containerized with Docker

This repo contains lecture notes and assignments of stanford's CS224N course of NLP

Language Generator Model. Can "Miguel de Cervantes" be digitally reincarnated?

solution for assessment at https://docs.google.com/forms/d/e/1FAIpQLSf4mqNST9kC6P41EDtaEYk65k0DptjuSyA_iRBoN10FwalSpw/viewform

Attempt at the Amazon ML Challenge 2024

An AI-powered Document Summarization System built with Streamlit and Python. It features an Extractive Track using statistical heuristics (TF-IDF/Frequency vectors) and an Abstractive Track utilizing deep learning (NLP Transformers) to process and synthesize long text documents efficiently. Created as part of the TEYZIX CORE Internship Program....