pop-ml 0.2.0

What is POP?

This project is built with pop, a Python-based implementation of Plugin Oriented Programming (POP). POP seeks to bring together concepts and wisdom from the history of computing in new ways to solve modern computing problems.

For more information:

• Intro to Plugin Oriented Programming (POP)

• pop-awesome

• pop-create

Prerequisites

• Python 3.8+

• git (if installing from source, or contributing to the project)

Installation

If wanting to use pop-ml, you can do so by either installing from PyPI or from source.

pip install pop-ml

git clone https://gitlab.com/vmware/pop/pop-ml.git

cd pop-ml

pip install .

CLI Examples

To use the pop-translate command-line tool, you can pass the text you want to translate as an argument, along with any additional options:

pop-translate "Hello, World!" --translate-to es

This command will translate the input text “Hello, world!” from English (en) to Spanish (es).

To see a full list of available options, run:

pop-translate --help

Python Examples

Here is an example of how to use pop-ml as a Python library:

import pop.hub

# Initialize the hub
hub = pop.hub.Hub()

# Add the "ml" dynamic namespace to the hub
hub.pop.sub.add(dyne_name="ml")

# Load config values onto hub.OPT
hub.pop.config.load(["pop_ml"], cli="pop_ml")

# Call the idempotent "init" of pop-ml's tokenizer using values from config
hub.ml.tokenizer.init(
    model_name=hub.OPT.pop_ml.model_name,
    dest_lang=hub.OPT.pop_ml.dest_lang,
    source_lang=hub.OPT.pop_ml.source_lang,
    pretrained_model=hub.OPT.pop_ml.pretrained_model_class,
    pretrained_tokenizer=hub.OPT.pop_ml.pretrained_tokenizer_class,
)
# Call the function to translate the text
result = hub.ml.tokenizer.translate([text])
print(result)

In this example, we initialize the hub, load the “ml” dynamic namespace and config values onto it, initialize the tokenizer, and call the function to translate the text. The output will be the translated text.