Datasets:

ArXiv:
File size: 7,757 Bytes
e1acc28
15203ea
 
e1acc28
15203ea
ca10b7a
 
 
 
 
88480b4
ba31342
ca10b7a
3f0bfaf
15203ea
ca10b7a
1232c97
ca10b7a
 
15203ea
ca10b7a
 
e1acc28
 
ca10b7a
e1acc28
 
 
ca10b7a
 
ba31342
ca10b7a
 
bd95969
3f0bfaf
e1acc28
15203ea
ca10b7a
15203ea
3f0bfaf
15203ea
ca10b7a
 
 
 
1232c97
723d3ae
15203ea
f1c2ac1
ca10b7a
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
e1acc28
 
 
 
 
 
 
ca10b7a
e1acc28
 
 
 
 
 
 
 
 
 
ca10b7a
e1acc28
ca10b7a
 
e1acc28
 
 
ca10b7a
 
 
 
 
 
 
 
 
e1acc28
 
 
ca10b7a
e1acc28
 
 
 
 
 
ca10b7a
 
 
e1acc28
 
 
 
 
 
 
 
ca10b7a
 
e1acc28
 
 
ca10b7a
723d3ae
ca10b7a
 
e1acc28
 
 
 
 
d53ac27
 
6e9b741
d53ac27
3f0bfaf
ca10b7a
 
e1acc28
ca10b7a
e1acc28
ca10b7a
 
 
 
 
 
e1acc28
 
 
ca10b7a
 
e1acc28
 
 
 
 
 
 
bd95969
 
e1acc28
1232c97
e1acc28
 
 
1232c97
ba31342
 
 
1232c97
ba31342
 
1232c97
 
 
ca10b7a
b123d50
e1acc28
 
ca10b7a
 
 
 
 
 
 
 
 
 
 
 
 
e1acc28
 
 
 
 
 
 
bd95969
e1acc28
 
bd95969
 
 
 
 
 
 
 
 
e1acc28
 
 
 
 
bd95969
e1acc28
 
 
 
 
 
 
 
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
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
from typing import Dict, Iterable, List

import evaluate
from datasets import Features, Value

from .artifact import __file__ as _
from .blocks import __file__ as _
from .card import __file__ as _
from .catalog import __file__ as _
from .collections import __file__ as _
from .dataclass import __file__ as _
from .dict_utils import __file__ as _
from .file_utils import __file__ as _
from .formats import __file__ as _
from .fusion import __file__ as _
from .generator_utils import __file__ as _
from .hf_utils import __file__ as _
from .instructions import __file__ as _
from .load import __file__ as _
from .loaders import __file__ as _
from .metrics import __file__ as _
from .normalizers import __file__ as _
from .operator import (MultiStreamOperator, SequentialOperator,
                       SequentialOperatorInitilizer, StreamInitializerOperator)
from .operator import __file__ as _
from .operators import (Apply, ApplyMetric, ApplyOperatorsField,
                        ApplyStreamOperatorsField, FlattenInstances,
                        MergeStreams, SplitByValue)
from .operators import __file__ as _
from .processors import __file__ as _
from .random_utils import __file__ as _
from .recipe import __file__ as _
from .register import __file__ as _
from .register import _reset_env_local_catalogs, register_all_artifacts
from .renderers import __file__ as _
from .schema import UNITXT_DATASET_SCHEMA
from .schema import __file__ as _
from .split_utils import __file__ as _
from .splitters import __file__ as _
from .standard import __file__ as _
from .stream import MultiStream, Stream
from .stream import __file__ as _
from .task import __file__ as _
from .templates import __file__ as _
from .text_utils import __file__ as _
from .type_utils import __file__ as _
from .utils import __file__ as _
from .validate import __file__ as _
from .version import __file__ as _


class MultiStreamScoreMean(MultiStreamOperator):
    def aggegate_results(self, multi_stream: MultiStream):
        scores = []
        for stream in multi_stream.values():
            instance = stream.peak()
            scores.append(instance["score"]["global"]["score"])

        from statistics import mean

        return mean(scores)

    def spread_results(self, stream: Stream, score: float):
        for instance in stream:
            instance["score"]["global"]["groups_mean_score"] = score
            yield instance

    def spread_results_one_stream(self, stream: Stream):
        for instance in stream:
            instance["score"]["global"]["groups_mean_score"] = instance["score"][
                "global"
            ]["score"]
            yield instance

    def process(self, multi_stream: MultiStream) -> MultiStream:
        result = {}

        # optimization in to avoid double calculation of metrics
        # when aggregating results, if there is only one stream.
        if len(multi_stream) == 1:
            for stream_name, stream in multi_stream.items():
                result[stream_name] = Stream(
                    self.spread_results_one_stream, gen_kwargs={"stream": stream}
                )
            return MultiStream(result)

        mean_score = self.aggegate_results(multi_stream)
        result = {}
        for stream_name, stream in multi_stream.items():
            result[stream_name] = Stream(
                self.spread_results, gen_kwargs={"stream": stream, "score": mean_score}
            )

        return MultiStream(result)


class FromPredictionsAndOriginalData(StreamInitializerOperator):
    def zip(self, predictions, references):
        for prediction, original in zip(predictions, references):
            yield {**original, "prediction": prediction}

    def process(
        self, predictions: List[str], references: Iterable, split_name: str = "all"
    ) -> MultiStream:
        return MultiStream(
            {
                split_name: Stream(
                    self.zip,
                    gen_kwargs={"predictions": predictions, "references": references},
                )
            }
        )


# The additional_inputs field in the schema is defined as
# Sequence({"key": Value(dtype="string"), "value": Value("string")})
# When receiving instances from this scheme, the keys and values are returned as two separate
# lists, and are converted to a dictionary.


def _from_key_value_pairs(key_value_list: Dict[str, list]) -> Dict[str, str]:
    return dict(zip(key_value_list["key"], key_value_list["value"]))


class MetricRecipe(SequentialOperatorInitilizer):
    calc_confidence_intervals: bool = True

    def prepare(self):
        register_all_artifacts()
        self.steps = [
            FromPredictionsAndOriginalData(),
            Apply(
                "additional_inputs",
                function=_from_key_value_pairs,
                to_field="additional_inputs",
            ),
            ApplyOperatorsField(
                inputs_fields=["prediction", "references"],
                fields_to_treat_as_list=["references"],
                operators_field="postprocessors",
                default_operators=["processors.to_string_stripped"],
            ),
            SplitByValue(["group"]),
            ApplyMetric(
                "metrics",
                calc_confidence_intervals=self.calc_confidence_intervals,
            ),
            MultiStreamScoreMean(),
            MergeStreams(),
        ]


UNITXT_METRIC_SCHEMA = Features(
    {"predictions": Value("string"), "references": dict(UNITXT_DATASET_SCHEMA)}
)


def _compute(
    predictions: List[str],
    references: Iterable,
    flatten: bool = False,
    split_name: str = "all",
    calc_confidence_intervals: bool = True,
):
    _reset_env_local_catalogs()
    register_all_artifacts()
    recipe = MetricRecipe(calc_confidence_intervals=calc_confidence_intervals)

    multi_stream = recipe(
        predictions=predictions, references=references, split_name=split_name
    )

    if flatten:
        operator = FlattenInstances()
        multi_stream = operator(multi_stream)

    stream = multi_stream[split_name]
    return list(stream)


# TODO: currently we have two classes with this name. metric.Metric and matrics.Metric...
# @evaluate.utils.file_utils.add_start_docstrings(_DESCRIPTION, _KWARGS_DESCRIPTION)
class Metric(evaluate.Metric):
    calc_confidence_intervals: bool = True

    def _info(self):
        return evaluate.MetricInfo(
            description="_DESCRIPTION",
            citation="_CITATION",
            # inputs_description=_KWARGS_DESCRIPTION,
            features=UNITXT_METRIC_SCHEMA,
            codebase_urls=["https://"],
            reference_urls=[
                "https://",
                "https://",
            ],
        )

    def _compute(
        self,
        predictions: List[str],
        references: Iterable,
        flatten: bool = False,
        split_name: str = "all",
    ):
        try:
            from unitxt.dataset import \
                get_dataset_artifact as get_dataset_artifact_installed

            unitxt_installed = True
        except ImportError:
            unitxt_installed = False

        if unitxt_installed:
            from unitxt.metric import _compute as _compute_installed

            return _compute_installed(
                predictions=predictions,
                references=references,
                flatten=flatten,
                split_name=split_name,
                calc_confidence_intervals=self.calc_confidence_intervals,
            )

        return _compute(
            predictions=predictions,
            references=references,
            flatten=flatten,
            split_name=split_name,
            calc_confidence_intervals=self.calc_confidence_intervals,
        )