A Deep-Learning-Based Persian Speech Recognition System

FreeCite: A Judge-Free Benchmark for Granular Citation Evaluation in Large Language Models

Transformers applied on Time Series Forecasting

Implementation of GRU-based Encoder-Decoder Architecture with Bahdanau Attention Mechanism for Machine Translation from German to English.

Attention in Seq2Seq with Teacher Forcing and Beam Search Decoding

Convolutional Language Model for Image Captioning

keras Neural Machine Translation using seq2seq with teacher forcing.

LSTM encoder-decoder model that translates infix mathematical expressions to postfix notation, achieving 100% accuracy with true autoregressive decoding. Built with TensorFlow/Keras.

Teacher-forced confidence analysis for language models. pip install confidence-cartography

Teacher-forced confidence as a false-belief sensor for language models.

A keras implementation of word level encoder-decoder architecture with teacher forcing.

This notebooks, we train a seq2seq decoder model with teacher forcing. Then use the trained layers from the decoder to generate a Sonnet.

Generate emotionally coherent song lyrics from melodies using deep learning. Combines LSTM-based language modeling with static and time-aligned MIDI features.

Interact with a trained chatbot that uses sequence to sequence model with luong attention mechanism over jointly trained encoder-decoder modules and implementation of greedy search decoding module.