microsoft/inmt

Interactive Neural Machine Translation tool

Interactive Neural Machine Translation

Assisting Translators with on-the-fly Translation Suggestions

Introduction

Interactive Machine Translation app uses Django and jQuery as its tech stack. Please refer to their docs for any doubts.

Installation Instructions

I. Prerequisites

1. Clone INMT and prepare MT models:

git clone https://github.com/microsoft/inmt

2. Make a new model folder using mkdir model where the models need to be placed. Models can be downloaded from here. These contain English to Hindi translation models in both directions.

3. If you want to train your own models, refer to Training MT Models.

Rest of the installation can be carried out either bare or using docker. Docker is preferable for its ease of installation.

II. Docker Installation

Assuming you have docker setup in your system, simply run docker-compose up -d. This application requires atleast 4GB of memory in order to run. Allot your docker memory accordingly.

III. Bare Installation

1. Install dependencies using - python -m pip install -r requirements.txt. Be sure to check your python version. This tool is compatible with Python3.
2. Install OpenNMT dependences using - cd opennmt & python setup.py install & cd -
3. Run the migrations and start the server - python manage.py makemigrations && python manage.py migrate && python manage.py runserver
4. The server opens on port 8000 by default. Open localhost:8000/simple for the simple interface.

Training MT Models

OpenNMT is used as the translation engine to power INMT. In order to train your own models, you need parallel sentences in your desired language. The basic instructions are listed as follows:

1. Go to opennmt folder: cd opennmt
2. Preprocess parallel (src & tgt) sentences: onmt_preprocess -train_src data/src-train.txt -train_tgt data/tgt-train.txt -valid_src data/src-val.txt -valid_tgt data/tgt-val.txt -save_data data/demo
3. Train your model (with GPUs): onmt_train -data data/demo -save_model demo-model

For advanced instructions on the training process, refer to OpenNMT docs.

Citation

If you find this work helpful, especially for research work, do consider citing us:

@inproceedings{santy-etal-2019-inmt,
    title = "{INMT}: Interactive Neural Machine Translation Prediction",
    author = "Santy, Sebastin  and
      Dandapat, Sandipan  and
      Choudhury, Monojit  and
      Bali, Kalika",
    booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP): System Demonstrations",
    month = nov,
    year = "2019",
    address = "Hong Kong, China",
    publisher = "Association for Computational Linguistics",
    url = "https://www.aclweb.org/anthology/D19-3018",
    doi = "10.18653/v1/D19-3018",
    pages = "103--108",
}

Contributing

This project welcomes contributions and suggestions. Most contributions require you to agree to a Contributor License Agreement (CLA) declaring that you have the right to, and actually do, grant us the rights to use your contribution. For details, visit https://cla.opensource.microsoft.com.

When you submit a pull request, a CLA bot will automatically determine whether you need to provide a CLA and decorate the PR appropriately (e.g., status check, comment). Simply follow the instructions provided by the bot. You will only need to do this once across all repos using our CLA.

This project has adopted the Microsoft Open Source Code of Conduct. For more information see the Code of Conduct FAQ or contact [email protected] with any additional questions or comments.

inmt's Issues

Set up Linting for the codebase

Add the lints to Azure Pipelines.

Implement Translation Memory (TM)

Currently, every time a translation is needed, the translation API is called and the request is processed for that source and partial input. Implementing Translation Memory would mean that we can use translation suggestions which were previously done by the translators.

Should the TM be user specific or global? It can probably be a combination of both - maybe batch and compare the individual TM and add that to global if it is commonly occurring.

Should the TM suggestion be shown with a different color explicitly to specify it comes from the TM whereas the others come from NMT?

Please update a new download model address, thanks.

Look at the question of the last person, do not know who to find the new model download address.

Add option for number of suggestions

Currently, the maximum number of suggestions in the drop-down box is 5. Can we add that as a parameter to the API?

Data folder?

Where can I get the parallel data to train this on?

Fix BPE pre/post processing

This includes:

• Source Side BPE tokens - This includes applying BPE code while translating.
• Target Side BPE tokens - This includes joining of suggestions.

Restructuring the results

Currently, the results from the API is in this form:

{
  "result": "Today 's weather is beautiful\nToday is \nToday 's \nThis day \nThe weather \nToday , ",
  "attn": [
    1,
    1,
    1,
    1
  ],
  "partial": "",
  "ppl": 3.230648083472593,
  "avg": -1.172682762145996
}

While this works for our case, it would be great to restructure it:

• "result" can be two lists - "full_sentence": [f1] and "part_sentence: [p1, p2, p3, ...]. The length of each list is controlled by #22
• "attn" can be restructured so that it makes sense independently. Maybe add the source sentence as well in the output?
• "ppl" and "avg" can be moved into a "metrics" section as both represent a way to measure the output.

Fix text pointer interaction

There are some issues with the text pointer interactions. The common ones observed are:

• If a translation box is left blank, the pointer behavior on preview tab is buggy.
• After clicking the preview and coming back, the pointer behavior is buggy.
• Reverse pointer does not work in the preview tab.

secret in file

this secret needs to be removed and rotated ASAP.

inmt/InteractiveTranslation/settings.py

Lines 23 to 24 in c8074bd

	# SECURITY WARNING: keep the secret key used in production secret!
	SECRET_KEY = '^@f24c3b-j&1z8l9^&8ut&bx7%2oz+hj%vt28g1k5dpre#t$5t'

Decouple translation engine configurations

Currently, the translation engine configurations are ingrained in the code, decouple the configurations to a separate json file and include in the .gitignore file.

FileNotFoundError when running "python manage.py"

Allow editing sentences in between

Currently, we are performing beam search based on the inputs at the end. However, there should be a method to change inputs in between, so that the suggestion can be made based on the prefix and suffix.

Relevant Literature:

1. Fast Lexically Constrained Decoding with Dynamic Beam Allocation for Neural Machine Translation
2. Lexically Constrained Decoding for Sequence Generation Using Grid Beam Search
3. Guided Open Vocabulary Image Captioning with Constrained Beam Search

Record keystrokes for different instances

Currently, recording of keystrokes get replaced every time the translator opens the translation interface.

TODOs:

• Add mechanism to append keystrokes rather than replacing them.
• Record interaction keystrokes - probably use xpath to record these.

Fix Transliteration Helper Interaction

Currently, the transliteration helping mechanism takes the english input and then retrieves the appropriate transliteration and shows to the user. However, there are interaction flaws with this:

• Showing up of English characters in the translation box, which overlaps with the hindi suggestion.

• Issues for users of Language Keyboards as even with that input, the transliterator calls the API. This can be fixed by limiting call to transliteration API only if the input unicode lies in the range of English (Latin) script.

i can't open the model download link

the page show that:
This link has been disabled.
Sorry, access through this link has been removed by admin policy. Please contact the person who shared it with you.

OpenNMT as a git submodule to this repository

Currently, we use a non-versioned copy of OpenNMT which becomes difficult to update when there is an update upstream. Figure out a way to update OpenNMT while keeping our wrappers intact.