-
Notifications
You must be signed in to change notification settings - Fork 388
/
squad.py
136 lines (110 loc) · 4.83 KB
/
squad.py
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
# Copyright The PyTorch Lightning team.
#
# 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.
from typing import Any, Callable, Dict, Optional
import torch
from torch import Tensor
from torchmetrics import Metric
from torchmetrics.functional.text.squad import (
PREDS_TYPE,
TARGETS_TYPE,
_squad_compute,
_squad_input_check,
_squad_update,
)
class SQuAD(Metric):
"""Calculate `SQuAD Metric`_ which corresponds to the scoring script for version 1 of the Stanford Question
Answering Dataset (SQuAD).
Args:
compute_on_step:
Forward only calls ``update()`` and returns None if this is set to False.
dist_sync_on_step:
Synchronize metric state across processes at each ``forward()``
before returning the value at the step.
process_group:
Specify the process group on which synchronization is called.
dist_sync_fn:
Callback that performs the allgather operation on the metric state. When `None`, DDP
will be used to perform the allgather.
Example:
>>> from torchmetrics import SQuAD
>>> preds = [{"prediction_text": "1976", "id": "56e10a3be3433e1400422b22"}]
>>> target = [{"answers": {"answer_start": [97], "text": ["1976"]}, "id": "56e10a3be3433e1400422b22"}]
>>> sqaud = SQuAD()
>>> sqaud(preds, target)
{'exact_match': tensor(100.), 'f1': tensor(100.)}
References:
[1] SQuAD: 100,000+ Questions for Machine Comprehension of Text by Pranav Rajpurkar, Jian Zhang, Konstantin
Lopyrev, Percy Liang `SQuAD Metric`_ .
"""
is_differentiable = False
higher_is_better = True
f1_score: Tensor
exact_match: Tensor
total: Tensor
def __init__(
self,
compute_on_step: bool = True,
dist_sync_on_step: bool = False,
process_group: Optional[Any] = None,
dist_sync_fn: Optional[Callable] = None,
):
super().__init__(
compute_on_step=compute_on_step,
dist_sync_on_step=dist_sync_on_step,
process_group=process_group,
dist_sync_fn=dist_sync_fn,
)
self.add_state(name="f1_score", default=torch.tensor(0, dtype=torch.float), dist_reduce_fx="sum")
self.add_state(name="exact_match", default=torch.tensor(0, dtype=torch.float), dist_reduce_fx="sum")
self.add_state(name="total", default=torch.tensor(0, dtype=torch.int), dist_reduce_fx="sum")
def update(self, preds: PREDS_TYPE, target: TARGETS_TYPE) -> None: # type: ignore
"""Compute F1 Score and Exact Match for a collection of predictions and references.
Args:
preds:
A Dictionary or List of Dictionary-s that map `id` and `prediction_text` to the respective values.
Example prediction:
.. code-block:: python
{"prediction_text": "TorchMetrics is awesome", "id": "123"}
target:
A Dictioinary or List of Dictionary-s that contain the `answers` and `id` in the SQuAD Format.
Example target:
.. code-block:: python
{
'answers': [{'answer_start': [1], 'text': ['This is a test answer']}],
'id': '1',
}
Reference SQuAD Format:
.. code-block:: python
{
'answers': {'answer_start': [1], 'text': ['This is a test text']},
'context': 'This is a test context.',
'id': '1',
'question': 'Is this a test?',
'title': 'train test'
}
Raises:
KeyError:
If the required keys are missing in either predictions or targets.
"""
preds_dict, target_dict = _squad_input_check(preds, target)
f1_score, exact_match, total = _squad_update(preds_dict, target_dict)
self.f1_score += f1_score
self.exact_match += exact_match
self.total += total
def compute(self) -> Dict[str, Tensor]:
"""Aggregate the F1 Score and Exact match for the batch.
Return:
Dictionary containing the F1 score, Exact match score for the batch.
"""
return _squad_compute(self.f1_score, self.exact_match, self.total)