LSTM_Seq

Build Your Own Simple Translator and Seq-to-Seq Translator

Overview This repository contains two implementations of a translator: Simple Translator: A basic implementation using rule-based or word-level translation techniques. Seq-to-Seq Translator: A deep learning-based model leveraging a sequence-to-sequence (Seq2Seq) architecture for language translation.

Features

Simple Translator:

1. Rule-based translation with customizable mappings.
2. Quick to set up for translating structured or predefined phrases.

Seq-to-Seq Translator

1. Neural network-based translation using encoder-decoder architectures.
2. Supports attention mechanisms for better context understanding.
3. Pretrained models can be fine-tuned with custom datasets.

Dataset

The dataset should be in parallel text format, with source and target sentences aligned line by line. This model can be trained on various text datasets. For this project, datasets in [languages] (e.g., English, Finnish, Italian, Latin, Punjabi) have been used. The data should be formatted in plain text files.

This repository contains jupyter noteboook.