triplet_margin_loss.py

# Copyright 2020 The HuggingFace Datasets Authors.
#
# 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.
"""Triplet Margin Loss metric."""

import datasets
import evaluate
import numpy as np


_DESCRIPTION = """
Triplet margin loss is a loss function that measures a relative similarity between the samples.\n
A triplet is comprised of reference input 'anchor (a)', matching input 'positive examples (p)' and non-matching input 'negative examples (n)'.
The loss function for each triplet is given by:\n
L(a, p, n) = max{d(a,p) - d(a,n) + margin, 0}\n
where d(x, y) is the 2nd order (Euclidean) pairwise distance between x and y.
"""


_KWARGS_DESCRIPTION = """
Args:
    anchor (`list` of `float`): Reference inputs.
    positive (`list` of `float`): Matching inputs.
    negative (`list` of `float`): Non-matching inputs.
    margin (`float`): Margin, default:`1.0`

Returns:
    triplet_margin_loss (`float`): Total loss.
Examples:
    Example 1-A simple example
        >>> triplet_margin_loss = evaluate.load("triplet_margin_loss")
        >>> results = triplet_margin_loss.compute(
            anchor=[-0.4765, 1.7133, 1.3971, -1.0121, 0.0732], 
            positive=[0.9218, 0.6305, 0.3381, 0.1412, 0.2607], 
            negative=[0.1971, 0.7246, 0.6729, 0.0941, 0.1011])
        >>> print(results)
        {'triplet_margin_loss': 1.59}
    Example 2-The same as Example 1, except with `margin` set to `2.0`.
        >>> triplet_margin_loss = evaluate.load("triplet_margin_loss")
        >>> results = triplet_margin_loss.compute(
            anchor=[-0.4765, 1.7133, 1.3971, -1.0121, 0.0732], 
            positive=[0.9218, 0.6305, 0.3381, 0.1412, 0.2607], 
            negative=[0.1971, 0.7246, 0.6729, 0.0941, 0.1011]), 
            margin=2.0)
        >>> print(results)
        {'triplet_margin_loss': 2.59}
"""


_CITATION = """
@article{scikit-learn,
  title={Scikit-learn: Machine Learning in {P}ython},
  author={Pedregosa, F. and Varoquaux, G. and Gramfort, A. and Michel, V.
         and Thirion, B. and Grisel, O. and Blondel, M. and Prettenhofer, P.
         and Weiss, R. and Dubourg, V. and Vanderplas, J. and Passos, A. and
         Cournapeau, D. and Brucher, M. and Perrot, M. and Duchesnay, E.},
  journal={Journal of Machine Learning Research},
  volume={12},
  pages={2825--2830},
  year={2011}
}
@article{schultz2003learning,
  title={Learning a distance metric from relative comparisons},
  author={Schultz, Matthew and Joachims, Thorsten},
  journal={Advances in neural information processing systems},
  volume={16},
  year={2003}
}
"""


@evaluate.utils.file_utils.add_start_docstrings(_DESCRIPTION, _KWARGS_DESCRIPTION)
class TripletMarginLoss(evaluate.EvaluationModule):
    def _info(self):
        return evaluate.EvaluationModuleInfo(
            description=_DESCRIPTION,
            citation=_CITATION,
            inputs_description=_KWARGS_DESCRIPTION,
            features=datasets.Features(
                {
                    "anchor": datasets.Sequence(datasets.Value("float", id="references")),
                    "positive": datasets.Sequence(datasets.Value("float"), id="sequence"),
                    "negative": datasets.Sequence(datasets.Value("float"), id="sequence"),
                    "margin": datasets.Value("float")
                }
            ),
            reference_urls=["https://proceedings.neurips.cc/paper/2003/hash/d3b1fb02964aa64e257f9f26a31f72cf-Abstract.html"],
        )

    def _compute(self, anchor, positive, negative, margin=1.0):
        d_a_p = np.sqrt(sum((anchor - positive)**2))
        d_a_n = np.sqrt(sum((anchor - negative)**2))
        return {
            "accuracy": float(
                max(d_a_p - d_a_n + margin, 0)
            )
        }