Skip to content

quinbez/Vanilla-Transformer

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

40 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Vanilla Transformer Model for Text Generation

Project Overview

This repository is dedicated to the development of a custom transformer model for text generation. The model utilizes the Penn Treebank dataset for training, aiming to produce coherent and meaningful text outputs.

Folder Structure

  • data_preprocessing: Contains scripts for processing raw text data, including tokenization, formatting, and preparation for model input.
  • model_architecture: Includes implementations of the transformer model, such as encoding layers, attention mechanisms, and feed-forward networks.
  • training_evaluation: Scripts for training the model, optimizing parameters, and evaluating performance using metrics like accuracy, loss, and perplexity.
  • tests: Scripts for testing Scripts of the whole project(different levels of testing).

Goals

  • Develop a transformer model from scratch based on the Penn Treebank dataset.
  • Experiment with different hyperparameters to optimize text generation quality.
  • Collaborate as a team using branches and pull requests to ensure smooth integration.

About

Building a custom transformer model for text generation.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages