File size: 5,412 Bytes
f93a3d8
 
 
 
 
 
 
 
 
 
 
 
 
bfa7fa8
f93a3d8
 
 
 
a45b5e7
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
f93a3d8
 
 
 
 
 
 
 
 
 
 
 
bfa7fa8
f93a3d8
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
97de291
 
f93a3d8
 
 
 
 
 
 
 
 
bfa7fa8
 
 
 
b768da6
a45b5e7
 
cb87d9f
 
f93a3d8
578bca9
bfa7fa8
 
97de291
bfa7fa8
97de291
bfa7fa8
97de291
bfa7fa8
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
# 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.
"""SEScore: a text generation evaluation metric"""

import evaluate
import datasets

import comet
from typing import Dict
import torch
from comet.encoders.base import Encoder
from comet.encoders.bert import BERTEncoder
from transformers import AutoModel, AutoTokenizer

class robertaEncoder(BERTEncoder):
    def __init__(self, pretrained_model: str) -> None:
        super(Encoder, self).__init__()
        self.tokenizer = AutoTokenizer.from_pretrained(pretrained_model)
        self.model = AutoModel.from_pretrained(
            pretrained_model, add_pooling_layer=False
        )
        self.model.encoder.output_hidden_states = True

    @classmethod
    def from_pretrained(cls, pretrained_model: str) -> Encoder:
        return robertaEncoder(pretrained_model)

    def forward(
        self, input_ids: torch.Tensor, attention_mask: torch.Tensor, **kwargs
    ) -> Dict[str, torch.Tensor]:
        last_hidden_states, _, all_layers = self.model(
            input_ids=input_ids,
            attention_mask=attention_mask,
            output_hidden_states=True,
            return_dict=False,
        )
        return {
            "sentemb": last_hidden_states[:, 0, :],
            "wordemb": last_hidden_states,
            "all_layers": all_layers,
            "attention_mask": attention_mask,
        }


# 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 = """\
SEScore is an evaluation metric that trys to compute an overall score to measure text generation quality.
"""


# 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}
"""

# TODO: Define external resources urls if needed
BAD_WORDS_URL = "http://url/to/external/resource/bad_words.txt"


@evaluate.utils.file_utils.add_start_docstrings(_DESCRIPTION, _KWARGS_DESCRIPTION)
class SEScore(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({
                'predictions': datasets.Value("string", id="sequence"),
                'references': datasets.Value("string", id="sequence"),
            }),
            # 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):
        """download SEScore checkpoints to compute the scores"""
        # Download SEScore checkpoint
        from comet import load_from_checkpoint
        import os
        from huggingface_hub import snapshot_download
        # initialize roberta into str2encoder
        comet.encoders.str2encoder['RoBERTa'] = robertaEncoder
        destination = snapshot_download(repo_id="xu1998hz/sescore_english_mt", revision="main")
        self.scorer = load_from_checkpoint(f'{destination}/checkpoint/sescore_english_mt.ckpt')

    def _compute(self, predictions, references, gpus=None, progress_bar=False):
        if gpus is None:
            gpus = 1 if torch.cuda.is_available() else 0

        data = {"src": references, "mt": predictions}
        print(data)
        data = [dict(zip(data, t)) for t in zip(*data.values())]
        print(data)
        scores, mean_score = self.scorer.predict(data, gpus=gpus, progress_bar=progress_bar)
        return {"mean_score": mean_score, "scores": scores}