blaser_2_0_qe.py

# Copyright 2020 The HuggingFace Datasets Authors and the current dataset script contributor.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
"""TODO: Add a description here."""

import evaluate
import datasets
import torch

from sonar.inference_pipelines.text import TextToEmbeddingModelPipeline
from sonar.models.blaser.loader import load_blaser_model


# TODO: Add BibTeX citation
_CITATION = """\
@InProceedings{huggingface:module,
title = {A great new module},
authors={huggingface, Inc.},
year={2020}
}
"""

# TODO: Add description of the module here
_DESCRIPTION = """\
This new module is designed to solve this great ML task and is crafted with a lot of care.
"""


# TODO: Add description of the arguments of the module here
_KWARGS_DESCRIPTION = """
Calculates how good are predictions given some references, using certain scores
Args:
    predictions: list of predictions to score. Each predictions
        should be a string with tokens separated by spaces.
    references: list of reference for each prediction. Each
        reference should be a string with tokens separated by spaces.
Returns:
    accuracy: description of the first score,
    another_score: description of the second score,
Examples:
    Examples should be written in doctest format, and should illustrate how
    to use the function.

    >>> my_new_module = evaluate.load("my_new_module")
    >>> results = my_new_module.compute(references=[0, 1], predictions=[0, 1])
    >>> print(results)
    {'accuracy': 1.0}
"""



@evaluate.utils.file_utils.add_start_docstrings(_DESCRIPTION, _KWARGS_DESCRIPTION)
class BLASER20QE(evaluate.Metric):
    """TODO: Short description of my evaluation module."""

    def _info(self):
        # TODO: Specifies the evaluate.EvaluationModuleInfo object
        return evaluate.MetricInfo(
            # This is the description that will appear on the modules page.
            module_type="metric",
            description=_DESCRIPTION,
            citation=_CITATION,
            inputs_description=_KWARGS_DESCRIPTION,
            # This defines the format of each prediction and reference
            features=datasets.Features({
                'sources': datasets.Value('string'),
                'translations': datasets.Value('string'),
            }),
            # See https://arrow.apache.org/docs/python/api/datatypes.html#factory-functions for the data types
            # Homepage of the module for documentation
            homepage="http://module.homepage",
            # Additional links to the codebase or references
            codebase_urls=["http://github.com/path/to/codebase/of/new_module"],
            reference_urls=["http://path.to.reference.url/new_module"]
        )

    def _download_and_prepare(self, dl_manager):
        """Optional: download external resources useful to compute the scores"""
        
        self.text_embedder = TextToEmbeddingModelPipeline(encoder="text_sonar_basic_encoder", tokenizer="text_sonar_basic_encoder")
        self.blaser_qe = load_blaser_model("blaser_2_0_qe").eval()
        #self.blaser_ref = load_blaser_model("blaser_2_0_ref").eval()

    def _compute(self, sources, translations):
        """Returns the scores"""
        # TODO: adjust the languages
        src_embs = self.text_embedder.predict(sources, source_lang="eng_Latn")
        mt_embs = self.text_embedder.predict(translations, source_lang="eng_Latn")

        with torch.inference_mode():
            #ref_score = blaser_ref(src=src_embs, ref=ref_embs, mt=mt_embs).mean().item()
            qe_score = self.blaser_qe(src=src_embs, mt=mt_embs).mean().item()

        return {
            "BLASER 2.0-QE": qe_score,
        }