Skip to main content

No project description provided

Project description

client.marian

A client for interacting with Marian-NMT websocket server (see MarianNMT).

How it wokrs

Grammatical Error Correction consists of 4 repositories: service.gec, library.gec, library.marian, 'client.marian:

  • service.gec (QAI based REST service) imports library.marian
  • library.marian calls service.marian via client.marian
  • service.marian (C++ websocket server), fed with sentences, corrects them

In case tomorrow a better model is released...

  • service.gec just instantize different Grammar object
  • library.marian just changes calls/tokenization steps to call ...
  • service.marian which is the main subject of change.

Usage

from marian_client import MarianClient

host = "0.0.0.0" 
port = "5000"

mc = MarianClient(PORT=port, HOST=host)

tokenized_sentence = "Alice like cats ."

success, corrected_sentence, error_info = mc(tokenized_sentence)

if success:
    print(corrected_sentence)
else:
    print(f"Call to MarianClient failed with error code {error_info[0]} and message {error_info[1]}")

Installation

pip install marian-client

License

Unlicensed.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

marian-client-0.11.0.tar.gz (3.5 kB view hashes)

Uploaded Source

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page