File size: 29,854 Bytes
1ce5e18
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
import inspect
import types
import warnings
from collections.abc import Iterable
from typing import TYPE_CHECKING, Dict, List, Optional, Tuple, Union

import numpy as np

from ..data import SquadExample, SquadFeatures, squad_convert_examples_to_features
from ..modelcard import ModelCard
from ..tokenization_utils import PreTrainedTokenizer
from ..utils import (
    PaddingStrategy,
    add_end_docstrings,
    is_tf_available,
    is_tokenizers_available,
    is_torch_available,
    logging,
)
from .base import PIPELINE_INIT_ARGS, ArgumentHandler, ChunkPipeline


logger = logging.get_logger(__name__)

if TYPE_CHECKING:
    from ..modeling_tf_utils import TFPreTrainedModel
    from ..modeling_utils import PreTrainedModel

    if is_tokenizers_available():
        import tokenizers

if is_tf_available():
    import tensorflow as tf

    from ..models.auto.modeling_tf_auto import TF_MODEL_FOR_QUESTION_ANSWERING_MAPPING_NAMES

    Dataset = None

if is_torch_available():
    import torch
    from torch.utils.data import Dataset

    from ..models.auto.modeling_auto import MODEL_FOR_QUESTION_ANSWERING_MAPPING_NAMES


def decode_spans(
    start: np.ndarray, end: np.ndarray, topk: int, max_answer_len: int, undesired_tokens: np.ndarray
) -> Tuple:
    """
    Take the output of any `ModelForQuestionAnswering` and will generate probabilities for each span to be the actual
    answer.

    In addition, it filters out some unwanted/impossible cases like answer len being greater than max_answer_len or
    answer end position being before the starting position. The method supports output the k-best answer through the
    topk argument.

    Args:
        start (`np.ndarray`): Individual start probabilities for each token.
        end (`np.ndarray`): Individual end probabilities for each token.
        topk (`int`): Indicates how many possible answer span(s) to extract from the model output.
        max_answer_len (`int`): Maximum size of the answer to extract from the model's output.
        undesired_tokens (`np.ndarray`): Mask determining tokens that can be part of the answer
    """
    # Ensure we have batch axis
    if start.ndim == 1:
        start = start[None]

    if end.ndim == 1:
        end = end[None]

    # Compute the score of each tuple(start, end) to be the real answer
    outer = np.matmul(np.expand_dims(start, -1), np.expand_dims(end, 1))

    # Remove candidate with end < start and end - start > max_answer_len
    candidates = np.tril(np.triu(outer), max_answer_len - 1)

    #  Inspired by Chen & al. (https://github.com/facebookresearch/DrQA)
    scores_flat = candidates.flatten()
    if topk == 1:
        idx_sort = [np.argmax(scores_flat)]
    elif len(scores_flat) < topk:
        idx_sort = np.argsort(-scores_flat)
    else:
        idx = np.argpartition(-scores_flat, topk)[0:topk]
        idx_sort = idx[np.argsort(-scores_flat[idx])]

    starts, ends = np.unravel_index(idx_sort, candidates.shape)[1:]
    desired_spans = np.isin(starts, undesired_tokens.nonzero()) & np.isin(ends, undesired_tokens.nonzero())
    starts = starts[desired_spans]
    ends = ends[desired_spans]
    scores = candidates[0, starts, ends]

    return starts, ends, scores


def select_starts_ends(
    start,
    end,
    p_mask,
    attention_mask,
    min_null_score=1000000,
    top_k=1,
    handle_impossible_answer=False,
    max_answer_len=15,
):
    """
    Takes the raw output of any `ModelForQuestionAnswering` and first normalizes its outputs and then uses
    `decode_spans()` to generate probabilities for each span to be the actual answer.

    Args:
        start (`np.ndarray`): Individual start logits for each token.
        end (`np.ndarray`): Individual end logits for each token.
        p_mask (`np.ndarray`): A mask with 1 for values that cannot be in the answer
        attention_mask (`np.ndarray`): The attention mask generated by the tokenizer
        min_null_score(`float`): The minimum null (empty) answer score seen so far.
        topk (`int`): Indicates how many possible answer span(s) to extract from the model output.
        handle_impossible_answer(`bool`): Whether to allow null (empty) answers
        max_answer_len (`int`): Maximum size of the answer to extract from the model's output.
    """
    # Ensure padded tokens & question tokens cannot belong to the set of candidate answers.
    undesired_tokens = np.abs(np.array(p_mask) - 1)

    if attention_mask is not None:
        undesired_tokens = undesired_tokens & attention_mask

    # Generate mask
    undesired_tokens_mask = undesired_tokens == 0.0

    # Make sure non-context indexes in the tensor cannot contribute to the softmax
    start = np.where(undesired_tokens_mask, -10000.0, start)
    end = np.where(undesired_tokens_mask, -10000.0, end)

    # Normalize logits and spans to retrieve the answer
    start = np.exp(start - start.max(axis=-1, keepdims=True))
    start = start / start.sum()

    end = np.exp(end - end.max(axis=-1, keepdims=True))
    end = end / end.sum()

    if handle_impossible_answer:
        min_null_score = min(min_null_score, (start[0, 0] * end[0, 0]).item())

    # Mask CLS
    start[0, 0] = end[0, 0] = 0.0

    starts, ends, scores = decode_spans(start, end, top_k, max_answer_len, undesired_tokens)
    return starts, ends, scores, min_null_score


class QuestionAnsweringArgumentHandler(ArgumentHandler):
    """
    QuestionAnsweringPipeline requires the user to provide multiple arguments (i.e. question & context) to be mapped to
    internal [`SquadExample`].

    QuestionAnsweringArgumentHandler manages all the possible to create a [`SquadExample`] from the command-line
    supplied arguments.
    """

    def normalize(self, item):
        if isinstance(item, SquadExample):
            return item
        elif isinstance(item, dict):
            for k in ["question", "context"]:
                if k not in item:
                    raise KeyError("You need to provide a dictionary with keys {question:..., context:...}")
                elif item[k] is None:
                    raise ValueError(f"`{k}` cannot be None")
                elif isinstance(item[k], str) and len(item[k]) == 0:
                    raise ValueError(f"`{k}` cannot be empty")

            return QuestionAnsweringPipeline.create_sample(**item)
        raise ValueError(f"{item} argument needs to be of type (SquadExample, dict)")

    def __call__(self, *args, **kwargs):
        # Detect where the actual inputs are
        if args is not None and len(args) > 0:
            if len(args) == 1:
                inputs = args[0]
            elif len(args) == 2 and {type(el) for el in args} == {str}:
                inputs = [{"question": args[0], "context": args[1]}]
            else:
                inputs = list(args)
        # Generic compatibility with sklearn and Keras
        # Batched data
        elif "X" in kwargs:
            inputs = kwargs["X"]
        elif "data" in kwargs:
            inputs = kwargs["data"]
        elif "question" in kwargs and "context" in kwargs:
            if isinstance(kwargs["question"], list) and isinstance(kwargs["context"], str):
                inputs = [{"question": Q, "context": kwargs["context"]} for Q in kwargs["question"]]
            elif isinstance(kwargs["question"], list) and isinstance(kwargs["context"], list):
                if len(kwargs["question"]) != len(kwargs["context"]):
                    raise ValueError("Questions and contexts don't have the same lengths")

                inputs = [{"question": Q, "context": C} for Q, C in zip(kwargs["question"], kwargs["context"])]
            elif isinstance(kwargs["question"], str) and isinstance(kwargs["context"], str):
                inputs = [{"question": kwargs["question"], "context": kwargs["context"]}]
            else:
                raise ValueError("Arguments can't be understood")
        else:
            raise ValueError(f"Unknown arguments {kwargs}")

        # When user is sending a generator we need to trust it's a valid example
        generator_types = (types.GeneratorType, Dataset) if Dataset is not None else (types.GeneratorType,)
        if isinstance(inputs, generator_types):
            return inputs

        # Normalize inputs
        if isinstance(inputs, dict):
            inputs = [inputs]
        elif isinstance(inputs, Iterable):
            # Copy to avoid overriding arguments
            inputs = list(inputs)
        else:
            raise ValueError(f"Invalid arguments {kwargs}")

        for i, item in enumerate(inputs):
            inputs[i] = self.normalize(item)

        return inputs


@add_end_docstrings(PIPELINE_INIT_ARGS)
class QuestionAnsweringPipeline(ChunkPipeline):
    """
    Question Answering pipeline using any `ModelForQuestionAnswering`. See the [question answering
    examples](../task_summary#question-answering) for more information.

    Example:

    ```python
    >>> from transformers import pipeline

    >>> oracle = pipeline(model="deepset/roberta-base-squad2")
    >>> oracle(question="Where do I live?", context="My name is Wolfgang and I live in Berlin")
    {'score': 0.9191, 'start': 34, 'end': 40, 'answer': 'Berlin'}
    ```

    Learn more about the basics of using a pipeline in the [pipeline tutorial](../pipeline_tutorial)

    This question answering pipeline can currently be loaded from [`pipeline`] using the following task identifier:
    `"question-answering"`.

    The models that this pipeline can use are models that have been fine-tuned on a question answering task. See the
    up-to-date list of available models on
    [huggingface.co/models](https://huggingface.co/models?filter=question-answering).
    """

    default_input_names = "question,context"
    handle_impossible_answer = False

    def __init__(
        self,
        model: Union["PreTrainedModel", "TFPreTrainedModel"],
        tokenizer: PreTrainedTokenizer,
        modelcard: Optional[ModelCard] = None,
        framework: Optional[str] = None,
        task: str = "",
        **kwargs,
    ):
        super().__init__(
            model=model,
            tokenizer=tokenizer,
            modelcard=modelcard,
            framework=framework,
            task=task,
            **kwargs,
        )

        self._args_parser = QuestionAnsweringArgumentHandler()
        self.check_model_type(
            TF_MODEL_FOR_QUESTION_ANSWERING_MAPPING_NAMES
            if self.framework == "tf"
            else MODEL_FOR_QUESTION_ANSWERING_MAPPING_NAMES
        )

    @staticmethod
    def create_sample(
        question: Union[str, List[str]], context: Union[str, List[str]]
    ) -> Union[SquadExample, List[SquadExample]]:
        """
        QuestionAnsweringPipeline leverages the [`SquadExample`] internally. This helper method encapsulate all the
        logic for converting question(s) and context(s) to [`SquadExample`].

        We currently support extractive question answering.

        Arguments:
            question (`str` or `List[str]`): The question(s) asked.
            context (`str` or `List[str]`): The context(s) in which we will look for the answer.

        Returns:
            One or a list of [`SquadExample`]: The corresponding [`SquadExample`] grouping question and context.
        """
        if isinstance(question, list):
            return [SquadExample(None, q, c, None, None, None) for q, c in zip(question, context)]
        else:
            return SquadExample(None, question, context, None, None, None)

    def _sanitize_parameters(
        self,
        padding=None,
        topk=None,
        top_k=None,
        doc_stride=None,
        max_answer_len=None,
        max_seq_len=None,
        max_question_len=None,
        handle_impossible_answer=None,
        align_to_words=None,
        **kwargs,
    ):
        # Set defaults values
        preprocess_params = {}
        if padding is not None:
            preprocess_params["padding"] = padding
        if doc_stride is not None:
            preprocess_params["doc_stride"] = doc_stride
        if max_question_len is not None:
            preprocess_params["max_question_len"] = max_question_len
        if max_seq_len is not None:
            preprocess_params["max_seq_len"] = max_seq_len

        postprocess_params = {}
        if topk is not None and top_k is None:
            warnings.warn("topk parameter is deprecated, use top_k instead", UserWarning)
            top_k = topk
        if top_k is not None:
            if top_k < 1:
                raise ValueError(f"top_k parameter should be >= 1 (got {top_k})")
            postprocess_params["top_k"] = top_k
        if max_answer_len is not None:
            if max_answer_len < 1:
                raise ValueError(f"max_answer_len parameter should be >= 1 (got {max_answer_len}")
        if max_answer_len is not None:
            postprocess_params["max_answer_len"] = max_answer_len
        if handle_impossible_answer is not None:
            postprocess_params["handle_impossible_answer"] = handle_impossible_answer
        if align_to_words is not None:
            postprocess_params["align_to_words"] = align_to_words
        return preprocess_params, {}, postprocess_params

    def __call__(self, *args, **kwargs):
        """
        Answer the question(s) given as inputs by using the context(s).

        Args:
            args ([`SquadExample`] or a list of [`SquadExample`]):
                One or several [`SquadExample`] containing the question and context.
            X ([`SquadExample`] or a list of [`SquadExample`], *optional*):
                One or several [`SquadExample`] containing the question and context (will be treated the same way as if
                passed as the first positional argument).
            data ([`SquadExample`] or a list of [`SquadExample`], *optional*):
                One or several [`SquadExample`] containing the question and context (will be treated the same way as if
                passed as the first positional argument).
            question (`str` or `List[str]`):
                One or several question(s) (must be used in conjunction with the `context` argument).
            context (`str` or `List[str]`):
                One or several context(s) associated with the question(s) (must be used in conjunction with the
                `question` argument).
            topk (`int`, *optional*, defaults to 1):
                The number of answers to return (will be chosen by order of likelihood). Note that we return less than
                topk answers if there are not enough options available within the context.
            doc_stride (`int`, *optional*, defaults to 128):
                If the context is too long to fit with the question for the model, it will be split in several chunks
                with some overlap. This argument controls the size of that overlap.
            max_answer_len (`int`, *optional*, defaults to 15):
                The maximum length of predicted answers (e.g., only answers with a shorter length are considered).
            max_seq_len (`int`, *optional*, defaults to 384):
                The maximum length of the total sentence (context + question) in tokens of each chunk passed to the
                model. The context will be split in several chunks (using `doc_stride` as overlap) if needed.
            max_question_len (`int`, *optional*, defaults to 64):
                The maximum length of the question after tokenization. It will be truncated if needed.
            handle_impossible_answer (`bool`, *optional*, defaults to `False`):
                Whether or not we accept impossible as an answer.
            align_to_words (`bool`, *optional*, defaults to `True`):
                Attempts to align the answer to real words. Improves quality on space separated langages. Might hurt on
                non-space-separated languages (like Japanese or Chinese)

        Return:
            A `dict` or a list of `dict`: Each result comes as a dictionary with the following keys:

            - **score** (`float`) -- The probability associated to the answer.
            - **start** (`int`) -- The character start index of the answer (in the tokenized version of the input).
            - **end** (`int`) -- The character end index of the answer (in the tokenized version of the input).
            - **answer** (`str`) -- The answer to the question.
        """

        # Convert inputs to features

        examples = self._args_parser(*args, **kwargs)
        if isinstance(examples, (list, tuple)) and len(examples) == 1:
            return super().__call__(examples[0], **kwargs)
        return super().__call__(examples, **kwargs)

    def preprocess(self, example, padding="do_not_pad", doc_stride=None, max_question_len=64, max_seq_len=None):
        # XXX: This is specal, args_parser will not handle anything generator or dataset like
        # For those we expect user to send a simple valid example either directly as a SquadExample or simple dict.
        # So we still need a little sanitation here.
        if isinstance(example, dict):
            example = SquadExample(None, example["question"], example["context"], None, None, None)

        if max_seq_len is None:
            max_seq_len = min(self.tokenizer.model_max_length, 384)
        if doc_stride is None:
            doc_stride = min(max_seq_len // 2, 128)

        if doc_stride > max_seq_len:
            raise ValueError(f"`doc_stride` ({doc_stride}) is larger than `max_seq_len` ({max_seq_len})")

        if not self.tokenizer.is_fast:
            features = squad_convert_examples_to_features(
                examples=[example],
                tokenizer=self.tokenizer,
                max_seq_length=max_seq_len,
                doc_stride=doc_stride,
                max_query_length=max_question_len,
                padding_strategy=PaddingStrategy.MAX_LENGTH,
                is_training=False,
                tqdm_enabled=False,
            )
        else:
            # Define the side we want to truncate / pad and the text/pair sorting
            question_first = self.tokenizer.padding_side == "right"

            encoded_inputs = self.tokenizer(
                text=example.question_text if question_first else example.context_text,
                text_pair=example.context_text if question_first else example.question_text,
                padding=padding,
                truncation="only_second" if question_first else "only_first",
                max_length=max_seq_len,
                stride=doc_stride,
                return_token_type_ids=True,
                return_overflowing_tokens=True,
                return_offsets_mapping=True,
                return_special_tokens_mask=True,
            )
            # When the input is too long, it's converted in a batch of inputs with overflowing tokens
            # and a stride of overlap between the inputs. If a batch of inputs is given, a special output
            # "overflow_to_sample_mapping" indicate which member of the encoded batch belong to which original batch sample.
            # Here we tokenize examples one-by-one so we don't need to use "overflow_to_sample_mapping".
            # "num_span" is the number of output samples generated from the overflowing tokens.
            num_spans = len(encoded_inputs["input_ids"])

            # p_mask: mask with 1 for token than cannot be in the answer (0 for token which can be in an answer)
            # We put 0 on the tokens from the context and 1 everywhere else (question and special tokens)
            p_mask = [
                [tok != 1 if question_first else 0 for tok in encoded_inputs.sequence_ids(span_id)]
                for span_id in range(num_spans)
            ]

            features = []
            for span_idx in range(num_spans):
                input_ids_span_idx = encoded_inputs["input_ids"][span_idx]
                attention_mask_span_idx = (
                    encoded_inputs["attention_mask"][span_idx] if "attention_mask" in encoded_inputs else None
                )
                token_type_ids_span_idx = (
                    encoded_inputs["token_type_ids"][span_idx] if "token_type_ids" in encoded_inputs else None
                )
                # keep the cls_token unmasked (some models use it to indicate unanswerable questions)
                if self.tokenizer.cls_token_id is not None:
                    cls_indices = np.nonzero(np.array(input_ids_span_idx) == self.tokenizer.cls_token_id)[0]
                    for cls_index in cls_indices:
                        p_mask[span_idx][cls_index] = 0
                submask = p_mask[span_idx]
                features.append(
                    SquadFeatures(
                        input_ids=input_ids_span_idx,
                        attention_mask=attention_mask_span_idx,
                        token_type_ids=token_type_ids_span_idx,
                        p_mask=submask,
                        encoding=encoded_inputs[span_idx],
                        # We don't use the rest of the values - and actually
                        # for Fast tokenizer we could totally avoid using SquadFeatures and SquadExample
                        cls_index=None,
                        token_to_orig_map={},
                        example_index=0,
                        unique_id=0,
                        paragraph_len=0,
                        token_is_max_context=0,
                        tokens=[],
                        start_position=0,
                        end_position=0,
                        is_impossible=False,
                        qas_id=None,
                    )
                )

        for i, feature in enumerate(features):
            fw_args = {}
            others = {}
            model_input_names = self.tokenizer.model_input_names + ["p_mask", "token_type_ids"]

            for k, v in feature.__dict__.items():
                if k in model_input_names:
                    if self.framework == "tf":
                        tensor = tf.constant(v)
                        if tensor.dtype == tf.int64:
                            tensor = tf.cast(tensor, tf.int32)
                        fw_args[k] = tf.expand_dims(tensor, 0)
                    elif self.framework == "pt":
                        tensor = torch.tensor(v)
                        if tensor.dtype == torch.int32:
                            tensor = tensor.long()
                        fw_args[k] = tensor.unsqueeze(0)
                else:
                    others[k] = v

            is_last = i == len(features) - 1
            yield {"example": example, "is_last": is_last, **fw_args, **others}

    def _forward(self, inputs):
        example = inputs["example"]
        model_inputs = {k: inputs[k] for k in self.tokenizer.model_input_names}
        # `XXXForSequenceClassification` models should not use `use_cache=True` even if it's supported
        model_forward = self.model.forward if self.framework == "pt" else self.model.call
        if "use_cache" in inspect.signature(model_forward).parameters.keys():
            model_inputs["use_cache"] = False
        output = self.model(**model_inputs)
        if isinstance(output, dict):
            return {"start": output["start_logits"], "end": output["end_logits"], "example": example, **inputs}
        else:
            start, end = output[:2]
            return {"start": start, "end": end, "example": example, **inputs}

    def postprocess(
        self,
        model_outputs,
        top_k=1,
        handle_impossible_answer=False,
        max_answer_len=15,
        align_to_words=True,
    ):
        min_null_score = 1000000  # large and positive
        answers = []
        for output in model_outputs:
            start_ = output["start"]
            end_ = output["end"]
            example = output["example"]
            p_mask = output["p_mask"]
            attention_mask = (
                output["attention_mask"].numpy() if output.get("attention_mask", None) is not None else None
            )

            starts, ends, scores, min_null_score = select_starts_ends(
                start_, end_, p_mask, attention_mask, min_null_score, top_k, handle_impossible_answer, max_answer_len
            )

            if not self.tokenizer.is_fast:
                char_to_word = np.array(example.char_to_word_offset)

                # Convert the answer (tokens) back to the original text
                # Score: score from the model
                # Start: Index of the first character of the answer in the context string
                # End: Index of the character following the last character of the answer in the context string
                # Answer: Plain text of the answer
                for s, e, score in zip(starts, ends, scores):
                    token_to_orig_map = output["token_to_orig_map"]
                    answers.append(
                        {
                            "score": score.item(),
                            "start": np.where(char_to_word == token_to_orig_map[s])[0][0].item(),
                            "end": np.where(char_to_word == token_to_orig_map[e])[0][-1].item(),
                            "answer": " ".join(example.doc_tokens[token_to_orig_map[s] : token_to_orig_map[e] + 1]),
                        }
                    )
            else:
                # Convert the answer (tokens) back to the original text
                # Score: score from the model
                # Start: Index of the first character of the answer in the context string
                # End: Index of the character following the last character of the answer in the context string
                # Answer: Plain text of the answer
                question_first = bool(self.tokenizer.padding_side == "right")
                enc = output["encoding"]

                # Encoding was *not* padded, input_ids *might*.
                # It doesn't make a difference unless we're padding on
                # the left hand side, since now we have different offsets
                # everywhere.
                if self.tokenizer.padding_side == "left":
                    offset = (output["input_ids"] == self.tokenizer.pad_token_id).numpy().sum()
                else:
                    offset = 0

                # Sometimes the max probability token is in the middle of a word so:
                # - we start by finding the right word containing the token with `token_to_word`
                # - then we convert this word in a character span with `word_to_chars`
                sequence_index = 1 if question_first else 0
                for s, e, score in zip(starts, ends, scores):
                    s = s - offset
                    e = e - offset

                    start_index, end_index = self.get_indices(enc, s, e, sequence_index, align_to_words)

                    answers.append(
                        {
                            "score": score.item(),
                            "start": start_index,
                            "end": end_index,
                            "answer": example.context_text[start_index:end_index],
                        }
                    )

        if handle_impossible_answer:
            answers.append({"score": min_null_score, "start": 0, "end": 0, "answer": ""})
        answers = sorted(answers, key=lambda x: x["score"], reverse=True)[:top_k]
        if len(answers) == 1:
            return answers[0]
        return answers

    def get_indices(
        self, enc: "tokenizers.Encoding", s: int, e: int, sequence_index: int, align_to_words: bool
    ) -> Tuple[int, int]:
        if align_to_words:
            try:
                start_word = enc.token_to_word(s)
                end_word = enc.token_to_word(e)
                start_index = enc.word_to_chars(start_word, sequence_index=sequence_index)[0]
                end_index = enc.word_to_chars(end_word, sequence_index=sequence_index)[1]
            except Exception:
                # Some tokenizers don't really handle words. Keep to offsets then.
                start_index = enc.offsets[s][0]
                end_index = enc.offsets[e][1]
        else:
            start_index = enc.offsets[s][0]
            end_index = enc.offsets[e][1]
        return start_index, end_index

    def span_to_answer(self, text: str, start: int, end: int) -> Dict[str, Union[str, int]]:
        """
        When decoding from token probabilities, this method maps token indexes to actual word in the initial context.

        Args:
            text (`str`): The actual context to extract the answer from.
            start (`int`): The answer starting token index.
            end (`int`): The answer end token index.

        Returns:
            Dictionary like `{'answer': str, 'start': int, 'end': int}`
        """
        words = []
        token_idx = char_start_idx = char_end_idx = chars_idx = 0

        for i, word in enumerate(text.split(" ")):
            token = self.tokenizer.tokenize(word)

            # Append words if they are in the span
            if start <= token_idx <= end:
                if token_idx == start:
                    char_start_idx = chars_idx

                if token_idx == end:
                    char_end_idx = chars_idx + len(word)

                words += [word]

            # Stop if we went over the end of the answer
            if token_idx > end:
                break

            # Append the subtokenization length to the running index
            token_idx += len(token)
            chars_idx += len(word) + 1

        # Join text with spaces
        return {
            "answer": " ".join(words),
            "start": max(0, char_start_idx),
            "end": min(len(text), char_end_idx),
        }