Neural machine translation between the writings of Shakespeare and modern English using TensorFlow

Last update: Dec 28, 2022

Overview

Shakespeare translations using TensorFlow

This is an example of using the new Google's TensorFlow library on monolingual translation going from modern English to Shakespeare based on research from Wei Xu.

Prepare

First download the TensorFlow library depending on your platform:

pip install https://storage.googleapis.com/tensorflow/mac/tensorflow-0.5.0-py2-none-any.whl # for mac
pip install https://storage.googleapis.com/tensorflow/linux/cpu/tensorflow-0.5.0-cp27-none-linux_x86_64.whl # for ubuntu

Grabs parallel data.
Gets train, dev split.
Builds vocabulary
Converts parallel data into ids

From the root directory:

python -m tensorshake.get_data
python -m tensorshake.prepare_corpus

Delete /cache to start anew.

Train

Use the example BASH script to train the model. This saves the check points in the --train_dir directory. If you run it again, the training process continues from the check point. To restart with fresh parameters, simply delete/rename the check points.

./run.sh

Results

Benchmarks from original paper. (Shakespeare -> Modern English)

Input	Output
i will bite thee by the ear for that jest .	i ’ ll bite you by the ear for that joke .
what further woe conspires against mine age ?	what ’ s true despair conspires against my old age ?
how doth my lady ?	how is my lady ?
hast thou slain tybalt ?	have you killed tybalt ?
an i might live to see thee married once , i have my wish .	if i could live to see you married, i ’ ve my wish .
benvolio , who began this bloody fray ?	benvolio , who started this bloody fight itself ?
what is your will ?	what do you want ?
call her forth to me .	bring her out to me .

Cherrypicked examples from this repo (Modern English -> Shakespeare)

Input	Output
but you’re not listening to me.	but you do not hear me .
Gregory, on my word, we will not be humiliated, like carrying coal.	regory , we 'll not carry coals .
but he got the promotion.	he is the friend .
i can hit quickly, if i'm motivated.	i strike , i am moved .
Did you just give us the finger, sir?	have you leave the thumb , sir ?
You don’t know what you’re doing!	you do not what you know you .
have you killed Tybalt?	hast thou slain tybalt ?
Why, Romeo, are you crazy?	why , art thou mad , mad ?

Pre-Trained Models

Here is a link for an example model: https://s3-us-west-2.amazonaws.com/foxtype-nlp/tensorshake/model_cache.zip

Possible improvements

word embeddings
beam search
language model reranking

Neural machine translation between the writings of Shakespeare and modern English using TensorFlow

Related tags

Overview

Shakespeare translations using TensorFlow

Prepare

Train

Results

Pre-Trained Models

Possible improvements

Owner

Motoki Wu

Semi-supevised Semantic Segmentation with High- and Low-level Consistency

Code for the paper "Adapting Monolingual Models: Data can be Scarce when Language Similarity is High"

PyGCL: A PyTorch Library for Graph Contrastive Learning

This is the official source code of "BiCAT: Bi-Chronological Augmentation of Transformer for Sequential Recommendation".

Pytorch implementation of Supporting Clustering with Contrastive Learning, NAACL 2021

AI4Good project for detecting waste in the environment

LOFO (Leave One Feature Out) Importance calculates the importances of a set of features based on a metric of choice,

Code for "Finding Regions of Heterogeneity in Decision-Making via Expected Conditional Covariance" at NeurIPS 2021

In this work, we will implement some basic but important algorithm of machine learning step by step.

Surrogate- and Invariance-Boosted Contrastive Learning (SIB-CL)

Code for AA-RMVSNet: Adaptive Aggregation Recurrent Multi-view Stereo Network (ICCV 2021).

Machine Learning Models were applied to predict the mass of the brain based on gender, age ranges, and head size.

OCTIS: Comparing Topic Models is Simple! A python package to optimize and evaluate topic models (accepted at EACL2021 demo track)

PyTorch implementation of UNet++ (Nested U-Net).

PRIME: A Few Primitives Can Boost Robustness to Common Corruptions

NLP From Scratch Without Large-Scale Pretraining: A Simple and Efficient Framework

A PyTorch Implementation of Single Shot Scale-invariant Face Detector.

DyNet: The Dynamic Neural Network Toolkit

This toolkit provides codes to download and pre-process the SLUE datasets, train the baseline models, and evaluate SLUE tasks.

Breaking the Curse of Space Explosion: Towards Efficient NAS with Curriculum Search