Spaces:
Runtime error
Runtime error
File size: 5,546 Bytes
54ad4a5 48fa02d 54ad4a5 d241401 54ad4a5 22fbc77 54ad4a5 e90329b 54ad4a5 e90329b 54ad4a5 22fbc77 54ad4a5 22fbc77 54ad4a5 48fa02d b037f19 54ad4a5 48fa02d 54ad4a5 e90329b 54ad4a5 |
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 |
# 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 importlib
import datasets
import evaluate
_CITATION = """\
@misc{ren2020codebleu,
title={CodeBLEU: a Method for Automatic Evaluation of Code Synthesis},
author={Shuo Ren and Daya Guo and Shuai Lu and Long Zhou and Shujie Liu and Duyu Tang and Neel Sundaresan and Ming Zhou and Ambrosio Blanco and Shuai Ma},
year={2020},
eprint={2009.10297},
archivePrefix={arXiv},
primaryClass={cs.SE}
}
"""
_DESCRIPTION = """\
Unofficial `CodeBLEU` implementation that supports Linux and MacOS.
"""
_KWARGS_DESCRIPTION = """
Calculate a weighted combination of `n-gram match (BLEU)`, `weighted n-gram match (BLEU-weighted)`, `AST match` and `data-flow match` 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.
language: programming language in ['java','js','c_sharp','php','c','python','cpp']. Please note that, due to the way Datasets works, the number of entities in the language array must match the number of entries in the predictions and references arrays, but only the first value from the languages array will be used. This means that you will not be able to compute a metric for different langauges at the same time, but mst do them as sequential calls to CodeBleu.
weights: tuple of 4 floats to use as weights for scores. Defaults to (0.25, 0.25, 0.25, 0.25).
Returns:
codebleu: resulting `CodeBLEU` score,
ngram_match_score: resulting `n-gram match (BLEU)` score,
weighted_ngram_match_score: resulting `weighted n-gram match (BLEU-weighted)` score,
syntax_match_score: resulting `AST match` score,
dataflow_match_score: resulting `data-flow match` score,
Examples:
>>> metric = evaluate.load("k4black/codebleu")
>>> ref = "def sum ( first , second ) :\n return second + first"
>>> pred = "def add ( a , b ) :\n return a + b"
>>> results = metric.compute(references=[ref], predictions=[pred], language=["python"])
>>> print(results)
"""
@evaluate.utils.file_utils.add_start_docstrings(_DESCRIPTION, _KWARGS_DESCRIPTION)
class codebleu(evaluate.Metric):
"""CodeBLEU metric from CodexGLUE"""
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.Sequence(datasets.Value("string", id="sequence"), id="references"),
"lang": datasets.Value("string"),
# "weights": datasets.Value("string"),
# "tokenizer": datasets.Value("string"),
}
),
datasets.Features(
{
"predictions": datasets.Value("string", id="sequence"),
"references": datasets.Value("string", id="sequence"),
"lang": datasets.Value("string"),
# "weights": datasets.Value("string"),
# "tokenizer": datasets.Value("string"),
}
),
],
# Homepage of the module for documentation
homepage="https://github.com/k4black/codebleu",
# Additional links to the codebase or references
codebase_urls=["https://github.com/k4black/codebleu"],
reference_urls=[
"https://github.com/k4black/codebleu",
"https://github.com/microsoft/CodeXGLUE/tree/main/Code-Code/code-to-code-trans/evaluator",
"https://arxiv.org/abs/2009.10297",
],
)
def _download_and_prepare(self, dl_manager):
"""Optional: download external resources useful to compute the scores"""
# workarounds as this file have to be named codebleu (evaluate library requirement)
self.codebleu_package = importlib.import_module("codebleu")
pass
def _compute(self, predictions, references, lang, weights=(0.25, 0.25, 0.25, 0.25), tokenizer=None):
"""Returns the scores"""
return self.codebleu_package.calc_codebleu(
references=references,
predictions=predictions,
lang=lang[0],
weights=weights,
tokenizer=tokenizer,
)
|