跳转至

luke

mindnlp.transformers.models.luke.modeling_luke

MindSpore LUKE model.

mindnlp.transformers.models.luke.modeling_luke.BaseLukeModelOutput dataclass

Bases: BaseModelOutput

Base class for model's outputs, with potential hidden states and attentions.

PARAMETER DESCRIPTION
last_hidden_state

Sequence of hidden-states at the output of the last layer of the model.

TYPE: `mindspore.Tensor` of shape `(batch_size, sequence_length, hidden_size)` DEFAULT: None

entity_last_hidden_state

Sequence of entity hidden-states at the output of the last layer of the model.

TYPE: `mindspore.Tensor` of shape `(batch_size, entity_length, hidden_size)` DEFAULT: None

hidden_states

Tuple of mindspore.Tensor (one for the output of the embeddings + one for the output of each layer) of shape (batch_size, sequence_length, hidden_size).

Hidden-states of the model at the output of each layer plus the initial embedding outputs.

TYPE: `tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True` DEFAULT: None

entity_hidden_states

Tuple of mindspore.Tensor (one for the output of the embeddings + one for the output of each layer) of shape (batch_size, entity_length, hidden_size). Entity hidden-states of the model at the output of each layer plus the initial entity embedding outputs.

TYPE: `tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True` DEFAULT: None

attentions

Tuple of mindspore.Tensor (one for each layer) of shape (batch_size, num_heads, sequence_length, sequence_length).

Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads.

TYPE: `tuple(mindspore.Tensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True` DEFAULT: None

Source code in mindnlp\transformers\models\luke\modeling_luke.py 
 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
@dataclass
class BaseLukeModelOutput(BaseModelOutput):
    """
    Base class for model's outputs, with potential hidden states and attentions.

    Args:
        last_hidden_state (`mindspore.Tensor` of shape `(batch_size, sequence_length, hidden_size)`):
            Sequence of hidden-states at the output of the last layer of the model.
        entity_last_hidden_state (`mindspore.Tensor` of shape `(batch_size, entity_length, hidden_size)`):
            Sequence of entity hidden-states at the output of the last layer of the model.
        hidden_states (`tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`):
            Tuple of `mindspore.Tensor` (one for the output of the embeddings + one for the output of each layer) of
            shape `(batch_size, sequence_length, hidden_size)`.

            Hidden-states of the model at the output of each layer plus the initial embedding outputs.
        entity_hidden_states (`tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`):
            Tuple of `mindspore.Tensor` (one for the output of the embeddings + one for the output of each layer) of
            shape `(batch_size, entity_length, hidden_size)`. Entity hidden-states of the model at the output of each
            layer plus the initial entity embedding outputs.
        attentions (`tuple(mindspore.Tensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`):
            Tuple of `mindspore.Tensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length,
            sequence_length)`.

            Attentions weights after the attention softmax, used to compute the weighted average in the self-attention
            heads.
    """

    entity_last_hidden_state: mindspore.Tensor = None
    entity_hidden_states: Optional[Tuple[mindspore.Tensor, ...]] = None

mindnlp.transformers.models.luke.modeling_luke.BaseLukeModelOutputWithPooling dataclass

Bases: BaseModelOutputWithPooling

Base class for outputs of the LUKE model.

PARAMETER DESCRIPTION
last_hidden_state

Sequence of hidden-states at the output of the last layer of the model.

TYPE: `mindspore.Tensor` of shape `(batch_size, sequence_length, hidden_size)` DEFAULT: None

entity_last_hidden_state

Sequence of entity hidden-states at the output of the last layer of the model.

TYPE: `mindspore.Tensor` of shape `(batch_size, entity_length, hidden_size)` DEFAULT: None

pooler_output

Last layer hidden-state of the first token of the sequence (classification token) further processed by a Linear layer and a Tanh activation function.

TYPE: `mindspore.Tensor` of shape `(batch_size, hidden_size)` DEFAULT: None

hidden_states

Tuple of mindspore.Tensor (one for the output of the embeddings + one for the output of each layer) of shape (batch_size, sequence_length, hidden_size). Hidden-states of the model at the output of each layer plus the initial embedding outputs.

TYPE: `tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True` DEFAULT: None

entity_hidden_states

Tuple of mindspore.Tensor (one for the output of the embeddings + one for the output of each layer) of shape (batch_size, entity_length, hidden_size). Entity hidden-states of the model at the output of each layer plus the initial entity embedding outputs.

TYPE: `tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True` DEFAULT: None

attentions

Tuple of mindspore.Tensor (one for each layer) of shape (batch_size, num_heads, sequence_length + entity_length, sequence_length + entity_length). Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads.

TYPE: `tuple(mindspore.Tensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True` DEFAULT: None

Source code in mindnlp\transformers\models\luke\modeling_luke.py 
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
@dataclass
class BaseLukeModelOutputWithPooling(BaseModelOutputWithPooling):
    """
    Base class for outputs of the LUKE model.

    Args:
        last_hidden_state (`mindspore.Tensor` of shape `(batch_size, sequence_length, hidden_size)`):
            Sequence of hidden-states at the output of the last layer of the model.
        entity_last_hidden_state (`mindspore.Tensor` of shape `(batch_size, entity_length, hidden_size)`):
            Sequence of entity hidden-states at the output of the last layer of the model.
        pooler_output (`mindspore.Tensor` of shape `(batch_size, hidden_size)`):
            Last layer hidden-state of the first token of the sequence (classification token) further processed by a
            Linear layer and a Tanh activation function.
        hidden_states (`tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`):
            Tuple of `mindspore.Tensor` (one for the output of the embeddings + one for the output of each layer) of
            shape `(batch_size, sequence_length, hidden_size)`. Hidden-states of the model at the output of each layer
            plus the initial embedding outputs.
        entity_hidden_states (`tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`):
            Tuple of `mindspore.Tensor` (one for the output of the embeddings + one for the output of each layer) of
            shape `(batch_size, entity_length, hidden_size)`. Entity hidden-states of the model at the output of each
            layer plus the initial entity embedding outputs.
        attentions (`tuple(mindspore.Tensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`):
            Tuple of `mindspore.Tensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length +
            entity_length, sequence_length + entity_length)`. Attentions weights after the attention softmax, used to
            compute the weighted average in the self-attention heads.
    """

    entity_last_hidden_state: mindspore.Tensor = None
    entity_hidden_states: Optional[Tuple[mindspore.Tensor, ...]] = None

mindnlp.transformers.models.luke.modeling_luke.EntityClassificationOutput dataclass

Bases: ModelOutput

Outputs of entity classification models.

PARAMETER DESCRIPTION
loss

Classification loss.

TYPE: `mindspore.Tensor` of shape `(1,)`, *optional*, returned when `labels` is provided DEFAULT: None

logits

Classification scores (before SoftMax).

TYPE: `mindspore.Tensor` of shape `(batch_size, config.num_labels)` DEFAULT: None

hidden_states

Tuple of mindspore.Tensor (one for the output of the embeddings + one for the output of each layer) of shape (batch_size, sequence_length, hidden_size). Hidden-states of the model at the output of each layer plus the initial embedding outputs.

TYPE: `tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True` DEFAULT: None

entity_hidden_states

Tuple of mindspore.Tensor (one for the output of the embeddings + one for the output of each layer) of shape (batch_size, entity_length, hidden_size). Entity hidden-states of the model at the output of each layer plus the initial entity embedding outputs.

TYPE: `tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True` DEFAULT: None

attentions

Tuple of mindspore.Tensor (one for each layer) of shape (batch_size, num_heads, sequence_length, sequence_length). Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads.

TYPE: `tuple(mindspore.Tensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True` DEFAULT: None

Source code in mindnlp\transformers\models\luke\modeling_luke.py 
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
@dataclass
class EntityClassificationOutput(ModelOutput):
    """
    Outputs of entity classification models.

    Args:
        loss (`mindspore.Tensor` of shape `(1,)`, *optional*, returned when `labels` is provided):
            Classification loss.
        logits (`mindspore.Tensor` of shape `(batch_size, config.num_labels)`):
            Classification scores (before SoftMax).
        hidden_states (`tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`):
            Tuple of `mindspore.Tensor` (one for the output of the embeddings + one for the output of each layer) of
            shape `(batch_size, sequence_length, hidden_size)`. Hidden-states of the model at the output of each layer
            plus the initial embedding outputs.
        entity_hidden_states (`tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`):
            Tuple of `mindspore.Tensor` (one for the output of the embeddings + one for the output of each layer) of
            shape `(batch_size, entity_length, hidden_size)`. Entity hidden-states of the model at the output of each
            layer plus the initial entity embedding outputs.
        attentions (`tuple(mindspore.Tensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`):
            Tuple of `mindspore.Tensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length,
            sequence_length)`. Attentions weights after the attention softmax, used to compute the weighted average in
            the self-attention heads.
    """

    loss: Optional[mindspore.Tensor] = None
    logits: mindspore.Tensor = None
    hidden_states: Optional[Tuple[mindspore.Tensor, ...]] = None
    entity_hidden_states: Optional[Tuple[mindspore.Tensor, ...]] = None
    attentions: Optional[Tuple[mindspore.Tensor, ...]] = None

mindnlp.transformers.models.luke.modeling_luke.EntityPairClassificationOutput dataclass

Bases: ModelOutput

Outputs of entity pair classification models.

PARAMETER DESCRIPTION
loss

Classification loss.

TYPE: `mindspore.Tensor` of shape `(1,)`, *optional*, returned when `labels` is provided DEFAULT: None

logits

Classification scores (before SoftMax).

TYPE: `mindspore.Tensor` of shape `(batch_size, config.num_labels)` DEFAULT: None

hidden_states

Tuple of mindspore.Tensor (one for the output of the embeddings + one for the output of each layer) of shape (batch_size, sequence_length, hidden_size). Hidden-states of the model at the output of each layer plus the initial embedding outputs.

TYPE: `tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True` DEFAULT: None

entity_hidden_states

Tuple of mindspore.Tensor (one for the output of the embeddings + one for the output of each layer) of shape (batch_size, entity_length, hidden_size). Entity hidden-states of the model at the output of each layer plus the initial entity embedding outputs.

TYPE: `tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True` DEFAULT: None

attentions

Tuple of mindspore.Tensor (one for each layer) of shape (batch_size, num_heads, sequence_length, sequence_length). Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads.

TYPE: `tuple(mindspore.Tensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True` DEFAULT: None

Source code in mindnlp\transformers\models\luke\modeling_luke.py 
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
@dataclass
class EntityPairClassificationOutput(ModelOutput):
    """
    Outputs of entity pair classification models.

    Args:
        loss (`mindspore.Tensor` of shape `(1,)`, *optional*, returned when `labels` is provided):
            Classification loss.
        logits (`mindspore.Tensor` of shape `(batch_size, config.num_labels)`):
            Classification scores (before SoftMax).
        hidden_states (`tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`):
            Tuple of `mindspore.Tensor` (one for the output of the embeddings + one for the output of each layer) of
            shape `(batch_size, sequence_length, hidden_size)`. Hidden-states of the model at the output of each layer
            plus the initial embedding outputs.
        entity_hidden_states (`tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`):
            Tuple of `mindspore.Tensor` (one for the output of the embeddings + one for the output of each layer) of
            shape `(batch_size, entity_length, hidden_size)`. Entity hidden-states of the model at the output of each
            layer plus the initial entity embedding outputs.
        attentions (`tuple(mindspore.Tensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`):
            Tuple of `mindspore.Tensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length,
            sequence_length)`. Attentions weights after the attention softmax, used to compute the weighted average in
            the self-attention heads.
    """

    loss: Optional[mindspore.Tensor] = None
    logits: mindspore.Tensor = None
    hidden_states: Optional[Tuple[mindspore.Tensor, ...]] = None
    entity_hidden_states: Optional[Tuple[mindspore.Tensor, ...]] = None
    attentions: Optional[Tuple[mindspore.Tensor, ...]] = None

mindnlp.transformers.models.luke.modeling_luke.EntitySpanClassificationOutput dataclass

Bases: ModelOutput

Outputs of entity span classification models.

PARAMETER DESCRIPTION
loss

Classification loss.

TYPE: `mindspore.Tensor` of shape `(1,)`, *optional*, returned when `labels` is provided DEFAULT: None

logits

Classification scores (before SoftMax).

TYPE: `mindspore.Tensor` of shape `(batch_size, entity_length, config.num_labels)` DEFAULT: None

hidden_states

Tuple of mindspore.Tensor (one for the output of the embeddings + one for the output of each layer) of shape (batch_size, sequence_length, hidden_size). Hidden-states of the model at the output of each layer plus the initial embedding outputs.

TYPE: `tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True` DEFAULT: None

entity_hidden_states

Tuple of mindspore.Tensor (one for the output of the embeddings + one for the output of each layer) of shape (batch_size, entity_length, hidden_size). Entity hidden-states of the model at the output of each layer plus the initial entity embedding outputs.

TYPE: `tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True` DEFAULT: None

attentions

Tuple of mindspore.Tensor (one for each layer) of shape (batch_size, num_heads, sequence_length, sequence_length). Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads.

TYPE: `tuple(mindspore.Tensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True` DEFAULT: None

Source code in mindnlp\transformers\models\luke\modeling_luke.py 
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
@dataclass
class EntitySpanClassificationOutput(ModelOutput):
    """
    Outputs of entity span classification models.

    Args:
        loss (`mindspore.Tensor` of shape `(1,)`, *optional*, returned when `labels` is provided):
            Classification loss.
        logits (`mindspore.Tensor` of shape `(batch_size, entity_length, config.num_labels)`):
            Classification scores (before SoftMax).
        hidden_states (`tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`):
            Tuple of `mindspore.Tensor` (one for the output of the embeddings + one for the output of each layer) of
            shape `(batch_size, sequence_length, hidden_size)`. Hidden-states of the model at the output of each layer
            plus the initial embedding outputs.
        entity_hidden_states (`tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`):
            Tuple of `mindspore.Tensor` (one for the output of the embeddings + one for the output of each layer) of
            shape `(batch_size, entity_length, hidden_size)`. Entity hidden-states of the model at the output of each
            layer plus the initial entity embedding outputs.
        attentions (`tuple(mindspore.Tensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`):
            Tuple of `mindspore.Tensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length,
            sequence_length)`. Attentions weights after the attention softmax, used to compute the weighted average in
            the self-attention heads.
    """

    loss: Optional[mindspore.Tensor] = None
    logits: mindspore.Tensor = None
    hidden_states: Optional[Tuple[mindspore.Tensor, ...]] = None
    entity_hidden_states: Optional[Tuple[mindspore.Tensor, ...]] = None
    attentions: Optional[Tuple[mindspore.Tensor, ...]] = None

mindnlp.transformers.models.luke.modeling_luke.LukeEmbeddings

Bases: Module

Same as BertEmbeddings with a tiny tweak for positional embeddings indexing.

Source code in mindnlp\transformers\models\luke\modeling_luke.py 
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
class LukeEmbeddings(nn.Module):
    """
    Same as BertEmbeddings with a tiny tweak for positional embeddings indexing.
    """

    def __init__(self, config):
        super().__init__()
        self.word_embeddings = nn.Embedding(config.vocab_size, config.hidden_size, padding_idx=config.pad_token_id)
        self.position_embeddings = nn.Embedding(config.max_position_embeddings, config.hidden_size)
        self.token_type_embeddings = nn.Embedding(config.type_vocab_size, config.hidden_size)

        # self.LayerNorm is not snake-cased to stick with TensorFlow model variable name and be able to load
        # any TensorFlow checkpoint file
        self.LayerNorm = nn.LayerNorm(config.hidden_size, eps=config.layer_norm_eps)
        self.dropout = nn.Dropout(config.hidden_dropout_prob)

        # End copy
        self.padding_idx = config.pad_token_id
        self.position_embeddings = nn.Embedding(
            config.max_position_embeddings, config.hidden_size, padding_idx=self.padding_idx
        )

    def forward(
        self,
        input_ids=None,
        token_type_ids=None,
        position_ids=None,
        inputs_embeds=None,
    ):
        if position_ids is None:
            if input_ids is not None:
                # Create the position ids from the input token ids. Any padded tokens remain padded.
                position_ids = create_position_ids_from_input_ids(input_ids, self.padding_idx)
            else:
                position_ids = self.create_position_ids_from_inputs_embeds(inputs_embeds)

        if input_ids is not None:
            input_shape = input_ids.shape
        else:
            input_shape = inputs_embeds.shape[:-1]

        if token_type_ids is None:
            token_type_ids = ops.zeros(input_shape, dtype=mindspore.int64)

        if inputs_embeds is None:
            inputs_embeds = self.word_embeddings(input_ids)

        position_embeddings = self.position_embeddings(position_ids)
        token_type_embeddings = self.token_type_embeddings(token_type_ids)

        embeddings = inputs_embeds + position_embeddings + token_type_embeddings
        embeddings = self.LayerNorm(embeddings)
        embeddings = self.dropout(embeddings)
        return embeddings

    def create_position_ids_from_inputs_embeds(self, inputs_embeds):
        """
        We are provided embeddings directly. We cannot infer which are padded so just generate sequential position ids.

        Args:
            inputs_embeds: mindspore.Tensor

        Returns: mindspore.Tensor
        """
        input_shape = inputs_embeds.shape[:-1]
        sequence_length = input_shape[1]

        position_ids = ops.arange(
            self.padding_idx + 1, sequence_length + self.padding_idx + 1, dtype=mindspore.int64
        )
        return position_ids.unsqueeze(0).broadcast_to(input_shape)

mindnlp.transformers.models.luke.modeling_luke.LukeEmbeddings.create_position_ids_from_inputs_embeds(inputs_embeds)

We are provided embeddings directly. We cannot infer which are padded so just generate sequential position ids.

PARAMETER DESCRIPTION
inputs_embeds

mindspore.Tensor
Source code in mindnlp\transformers\models\luke\modeling_luke.py 
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
def create_position_ids_from_inputs_embeds(self, inputs_embeds):
    """
    We are provided embeddings directly. We cannot infer which are padded so just generate sequential position ids.

    Args:
        inputs_embeds: mindspore.Tensor

    Returns: mindspore.Tensor
    """
    input_shape = inputs_embeds.shape[:-1]
    sequence_length = input_shape[1]

    position_ids = ops.arange(
        self.padding_idx + 1, sequence_length + self.padding_idx + 1, dtype=mindspore.int64
    )
    return position_ids.unsqueeze(0).broadcast_to(input_shape)

mindnlp.transformers.models.luke.modeling_luke.LukeForEntityClassification

Bases: LukePreTrainedModel

Source code in mindnlp\transformers\models\luke\modeling_luke.py 
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
class LukeForEntityClassification(LukePreTrainedModel):
    def __init__(self, config):
        super().__init__(config)

        self.luke = LukeModel(config)

        self.num_labels = config.num_labels
        self.dropout = nn.Dropout(config.hidden_dropout_prob)
        self.classifier = nn.Linear(config.hidden_size, config.num_labels)

        # Initialize weights and apply final processing
        self.post_init()

    def forward(
        self,
        input_ids: Optional[mindspore.Tensor] = None,
        attention_mask: Optional[mindspore.Tensor] = None,
        token_type_ids: Optional[mindspore.Tensor] = None,
        position_ids: Optional[mindspore.Tensor] = None,
        entity_ids: Optional[mindspore.Tensor] = None,
        entity_attention_mask: Optional[mindspore.Tensor] = None,
        entity_token_type_ids: Optional[mindspore.Tensor] = None,
        entity_position_ids: Optional[mindspore.Tensor] = None,
        head_mask: Optional[mindspore.Tensor] = None,
        inputs_embeds: Optional[mindspore.Tensor] = None,
        labels: Optional[mindspore.Tensor] = None,
        output_attentions: Optional[bool] = None,
        output_hidden_states: Optional[bool] = None,
        return_dict: Optional[bool] = None,
    ) -> Union[Tuple, EntityClassificationOutput]:
        r"""
        labels (`mindspore.Tensor` of shape `(batch_size,)` or `(batch_size, num_labels)`, *optional*):
            Labels for computing the classification loss. If the shape is `(batch_size,)`, the cross entropy loss is
            used for the single-label classification. In this case, labels should contain the indices that should be in
            `[0, ..., config.num_labels - 1]`. If the shape is `(batch_size, num_labels)`, the binary cross entropy
            loss is used for the multi-label classification. In this case, labels should only contain `[0, 1]`, where 0
            and 1 indicate false and true, respectively.

        Returns:

        Examples:

        ```python
        >>> from transformers import AutoTokenizer, LukeForEntityClassification

        >>> tokenizer = AutoTokenizer.from_pretrained("studio-ousia/luke-large-finetuned-open-entity")
        >>> model = LukeForEntityClassification.from_pretrained("studio-ousia/luke-large-finetuned-open-entity")

        >>> text = "Beyoncé lives in Los Angeles."
        >>> entity_spans = [(0, 7)]  # character-based entity span corresponding to "Beyoncé"
        >>> inputs = tokenizer(text, entity_spans=entity_spans, return_tensors="ms")
        >>> outputs = model(**inputs)
        >>> logits = outputs.logits
        >>> predicted_class_idx = logits.argmax(-1).item()
        >>> print("Predicted class:", model.config.id2label[predicted_class_idx])
        Predicted class: person
        ```"""
        return_dict = return_dict if return_dict is not None else self.config.use_return_dict

        outputs = self.luke(
            input_ids=input_ids,
            attention_mask=attention_mask,
            token_type_ids=token_type_ids,
            position_ids=position_ids,
            entity_ids=entity_ids,
            entity_attention_mask=entity_attention_mask,
            entity_token_type_ids=entity_token_type_ids,
            entity_position_ids=entity_position_ids,
            head_mask=head_mask,
            inputs_embeds=inputs_embeds,
            output_attentions=output_attentions,
            output_hidden_states=output_hidden_states,
            return_dict=True,
        )

        feature_vector = outputs.entity_last_hidden_state[:, 0, :]
        feature_vector = self.dropout(feature_vector)
        logits = self.classifier(feature_vector)

        loss = None
        if labels is not None:
            # When the number of dimension of `labels` is 1, cross entropy is used as the loss function. The binary
            # cross entropy is used otherwise.
            if labels.ndim == 1:
                loss = nn.functional.cross_entropy(logits, labels)
            else:
                loss = nn.functional.binary_cross_entropy_with_logits(logits.view(-1), labels.view(-1).type_as(logits))

        if not return_dict:
            return tuple(
                v
                for v in [loss, logits, outputs.hidden_states, outputs.entity_hidden_states, outputs.attentions]
                if v is not None
            )

        return EntityClassificationOutput(
            loss=loss,
            logits=logits,
            hidden_states=outputs.hidden_states,
            entity_hidden_states=outputs.entity_hidden_states,
            attentions=outputs.attentions,
        )

mindnlp.transformers.models.luke.modeling_luke.LukeForEntityClassification.forward(input_ids=None, attention_mask=None, token_type_ids=None, position_ids=None, entity_ids=None, entity_attention_mask=None, entity_token_type_ids=None, entity_position_ids=None, head_mask=None, inputs_embeds=None, labels=None, output_attentions=None, output_hidden_states=None, return_dict=None)

labels (mindspore.Tensor of shape (batch_size,) or (batch_size, num_labels), optional): Labels for computing the classification loss. If the shape is (batch_size,), the cross entropy loss is used for the single-label classification. In this case, labels should contain the indices that should be in [0, ..., config.num_labels - 1]. If the shape is (batch_size, num_labels), the binary cross entropy loss is used for the multi-label classification. In this case, labels should only contain [0, 1], where 0 and 1 indicate false and true, respectively.

Returns:

Examples:

>>> from transformers import AutoTokenizer, LukeForEntityClassification

>>> tokenizer = AutoTokenizer.from_pretrained("studio-ousia/luke-large-finetuned-open-entity")
>>> model = LukeForEntityClassification.from_pretrained("studio-ousia/luke-large-finetuned-open-entity")

>>> text = "Beyoncé lives in Los Angeles."
>>> entity_spans = [(0, 7)]  # character-based entity span corresponding to "Beyoncé"
>>> inputs = tokenizer(text, entity_spans=entity_spans, return_tensors="ms")
>>> outputs = model(**inputs)
>>> logits = outputs.logits
>>> predicted_class_idx = logits.argmax(-1).item()
>>> print("Predicted class:", model.config.id2label[predicted_class_idx])
Predicted class: person
Source code in mindnlp\transformers\models\luke\modeling_luke.py 
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
def forward(
    self,
    input_ids: Optional[mindspore.Tensor] = None,
    attention_mask: Optional[mindspore.Tensor] = None,
    token_type_ids: Optional[mindspore.Tensor] = None,
    position_ids: Optional[mindspore.Tensor] = None,
    entity_ids: Optional[mindspore.Tensor] = None,
    entity_attention_mask: Optional[mindspore.Tensor] = None,
    entity_token_type_ids: Optional[mindspore.Tensor] = None,
    entity_position_ids: Optional[mindspore.Tensor] = None,
    head_mask: Optional[mindspore.Tensor] = None,
    inputs_embeds: Optional[mindspore.Tensor] = None,
    labels: Optional[mindspore.Tensor] = None,
    output_attentions: Optional[bool] = None,
    output_hidden_states: Optional[bool] = None,
    return_dict: Optional[bool] = None,
) -> Union[Tuple, EntityClassificationOutput]:
    r"""
    labels (`mindspore.Tensor` of shape `(batch_size,)` or `(batch_size, num_labels)`, *optional*):
        Labels for computing the classification loss. If the shape is `(batch_size,)`, the cross entropy loss is
        used for the single-label classification. In this case, labels should contain the indices that should be in
        `[0, ..., config.num_labels - 1]`. If the shape is `(batch_size, num_labels)`, the binary cross entropy
        loss is used for the multi-label classification. In this case, labels should only contain `[0, 1]`, where 0
        and 1 indicate false and true, respectively.

    Returns:

    Examples:

    ```python
    >>> from transformers import AutoTokenizer, LukeForEntityClassification

    >>> tokenizer = AutoTokenizer.from_pretrained("studio-ousia/luke-large-finetuned-open-entity")
    >>> model = LukeForEntityClassification.from_pretrained("studio-ousia/luke-large-finetuned-open-entity")

    >>> text = "Beyoncé lives in Los Angeles."
    >>> entity_spans = [(0, 7)]  # character-based entity span corresponding to "Beyoncé"
    >>> inputs = tokenizer(text, entity_spans=entity_spans, return_tensors="ms")
    >>> outputs = model(**inputs)
    >>> logits = outputs.logits
    >>> predicted_class_idx = logits.argmax(-1).item()
    >>> print("Predicted class:", model.config.id2label[predicted_class_idx])
    Predicted class: person
    ```"""
    return_dict = return_dict if return_dict is not None else self.config.use_return_dict

    outputs = self.luke(
        input_ids=input_ids,
        attention_mask=attention_mask,
        token_type_ids=token_type_ids,
        position_ids=position_ids,
        entity_ids=entity_ids,
        entity_attention_mask=entity_attention_mask,
        entity_token_type_ids=entity_token_type_ids,
        entity_position_ids=entity_position_ids,
        head_mask=head_mask,
        inputs_embeds=inputs_embeds,
        output_attentions=output_attentions,
        output_hidden_states=output_hidden_states,
        return_dict=True,
    )

    feature_vector = outputs.entity_last_hidden_state[:, 0, :]
    feature_vector = self.dropout(feature_vector)
    logits = self.classifier(feature_vector)

    loss = None
    if labels is not None:
        # When the number of dimension of `labels` is 1, cross entropy is used as the loss function. The binary
        # cross entropy is used otherwise.
        if labels.ndim == 1:
            loss = nn.functional.cross_entropy(logits, labels)
        else:
            loss = nn.functional.binary_cross_entropy_with_logits(logits.view(-1), labels.view(-1).type_as(logits))

    if not return_dict:
        return tuple(
            v
            for v in [loss, logits, outputs.hidden_states, outputs.entity_hidden_states, outputs.attentions]
            if v is not None
        )

    return EntityClassificationOutput(
        loss=loss,
        logits=logits,
        hidden_states=outputs.hidden_states,
        entity_hidden_states=outputs.entity_hidden_states,
        attentions=outputs.attentions,
    )

mindnlp.transformers.models.luke.modeling_luke.LukeForEntityPairClassification

Bases: LukePreTrainedModel

Source code in mindnlp\transformers\models\luke\modeling_luke.py 
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
class LukeForEntityPairClassification(LukePreTrainedModel):
    def __init__(self, config):
        super().__init__(config)

        self.luke = LukeModel(config)

        self.num_labels = config.num_labels
        self.dropout = nn.Dropout(config.hidden_dropout_prob)
        self.classifier = nn.Linear(config.hidden_size * 2, config.num_labels, False)

        # Initialize weights and apply final processing
        self.post_init()

    def forward(
        self,
        input_ids: Optional[mindspore.Tensor] = None,
        attention_mask: Optional[mindspore.Tensor] = None,
        token_type_ids: Optional[mindspore.Tensor] = None,
        position_ids: Optional[mindspore.Tensor] = None,
        entity_ids: Optional[mindspore.Tensor] = None,
        entity_attention_mask: Optional[mindspore.Tensor] = None,
        entity_token_type_ids: Optional[mindspore.Tensor] = None,
        entity_position_ids: Optional[mindspore.Tensor] = None,
        head_mask: Optional[mindspore.Tensor] = None,
        inputs_embeds: Optional[mindspore.Tensor] = None,
        labels: Optional[mindspore.Tensor] = None,
        output_attentions: Optional[bool] = None,
        output_hidden_states: Optional[bool] = None,
        return_dict: Optional[bool] = None,
    ) -> Union[Tuple, EntityPairClassificationOutput]:
        r"""
        labels (`mindspore.Tensor` of shape `(batch_size,)` or `(batch_size, num_labels)`, *optional*):
            Labels for computing the classification loss. If the shape is `(batch_size,)`, the cross entropy loss is
            used for the single-label classification. In this case, labels should contain the indices that should be in
            `[0, ..., config.num_labels - 1]`. If the shape is `(batch_size, num_labels)`, the binary cross entropy
            loss is used for the multi-label classification. In this case, labels should only contain `[0, 1]`, where 0
            and 1 indicate false and true, respectively.

        Returns:

        Examples:

        ```python
        >>> from transformers import AutoTokenizer, LukeForEntityPairClassification

        >>> tokenizer = AutoTokenizer.from_pretrained("studio-ousia/luke-large-finetuned-tacred")
        >>> model = LukeForEntityPairClassification.from_pretrained("studio-ousia/luke-large-finetuned-tacred")

        >>> text = "Beyoncé lives in Los Angeles."
        >>> entity_spans = [
        ...     (0, 7),
        ...     (17, 28),
        ... ]  # character-based entity spans corresponding to "Beyoncé" and "Los Angeles"
        >>> inputs = tokenizer(text, entity_spans=entity_spans, return_tensors="ms")
        >>> outputs = model(**inputs)
        >>> logits = outputs.logits
        >>> predicted_class_idx = logits.argmax(-1).item()
        >>> print("Predicted class:", model.config.id2label[predicted_class_idx])
        Predicted class: per:cities_of_residence
        ```"""
        return_dict = return_dict if return_dict is not None else self.config.use_return_dict

        outputs = self.luke(
            input_ids=input_ids,
            attention_mask=attention_mask,
            token_type_ids=token_type_ids,
            position_ids=position_ids,
            entity_ids=entity_ids,
            entity_attention_mask=entity_attention_mask,
            entity_token_type_ids=entity_token_type_ids,
            entity_position_ids=entity_position_ids,
            head_mask=head_mask,
            inputs_embeds=inputs_embeds,
            output_attentions=output_attentions,
            output_hidden_states=output_hidden_states,
            return_dict=True,
        )

        feature_vector = ops.cat(
            [outputs.entity_last_hidden_state[:, 0, :], outputs.entity_last_hidden_state[:, 1, :]], dim=1
        )
        feature_vector = self.dropout(feature_vector)
        logits = self.classifier(feature_vector)

        loss = None
        if labels is not None:
            # When the number of dimension of `labels` is 1, cross entropy is used as the loss function. The binary
            # cross entropy is used otherwise.
            if labels.ndim == 1:
                loss = nn.functional.cross_entropy(logits, labels)
            else:
                loss = nn.functional.binary_cross_entropy_with_logits(logits.view(-1), labels.view(-1).type_as(logits))

        if not return_dict:
            return tuple(
                v
                for v in [loss, logits, outputs.hidden_states, outputs.entity_hidden_states, outputs.attentions]
                if v is not None
            )

        return EntityPairClassificationOutput(
            loss=loss,
            logits=logits,
            hidden_states=outputs.hidden_states,
            entity_hidden_states=outputs.entity_hidden_states,
            attentions=outputs.attentions,
        )

mindnlp.transformers.models.luke.modeling_luke.LukeForEntityPairClassification.forward(input_ids=None, attention_mask=None, token_type_ids=None, position_ids=None, entity_ids=None, entity_attention_mask=None, entity_token_type_ids=None, entity_position_ids=None, head_mask=None, inputs_embeds=None, labels=None, output_attentions=None, output_hidden_states=None, return_dict=None)

labels (mindspore.Tensor of shape (batch_size,) or (batch_size, num_labels), optional): Labels for computing the classification loss. If the shape is (batch_size,), the cross entropy loss is used for the single-label classification. In this case, labels should contain the indices that should be in [0, ..., config.num_labels - 1]. If the shape is (batch_size, num_labels), the binary cross entropy loss is used for the multi-label classification. In this case, labels should only contain [0, 1], where 0 and 1 indicate false and true, respectively.

Returns:

Examples:

>>> from transformers import AutoTokenizer, LukeForEntityPairClassification

>>> tokenizer = AutoTokenizer.from_pretrained("studio-ousia/luke-large-finetuned-tacred")
>>> model = LukeForEntityPairClassification.from_pretrained("studio-ousia/luke-large-finetuned-tacred")

>>> text = "Beyoncé lives in Los Angeles."
>>> entity_spans = [
...     (0, 7),
...     (17, 28),
... ]  # character-based entity spans corresponding to "Beyoncé" and "Los Angeles"
>>> inputs = tokenizer(text, entity_spans=entity_spans, return_tensors="ms")
>>> outputs = model(**inputs)
>>> logits = outputs.logits
>>> predicted_class_idx = logits.argmax(-1).item()
>>> print("Predicted class:", model.config.id2label[predicted_class_idx])
Predicted class: per:cities_of_residence
Source code in mindnlp\transformers\models\luke\modeling_luke.py 
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
def forward(
    self,
    input_ids: Optional[mindspore.Tensor] = None,
    attention_mask: Optional[mindspore.Tensor] = None,
    token_type_ids: Optional[mindspore.Tensor] = None,
    position_ids: Optional[mindspore.Tensor] = None,
    entity_ids: Optional[mindspore.Tensor] = None,
    entity_attention_mask: Optional[mindspore.Tensor] = None,
    entity_token_type_ids: Optional[mindspore.Tensor] = None,
    entity_position_ids: Optional[mindspore.Tensor] = None,
    head_mask: Optional[mindspore.Tensor] = None,
    inputs_embeds: Optional[mindspore.Tensor] = None,
    labels: Optional[mindspore.Tensor] = None,
    output_attentions: Optional[bool] = None,
    output_hidden_states: Optional[bool] = None,
    return_dict: Optional[bool] = None,
) -> Union[Tuple, EntityPairClassificationOutput]:
    r"""
    labels (`mindspore.Tensor` of shape `(batch_size,)` or `(batch_size, num_labels)`, *optional*):
        Labels for computing the classification loss. If the shape is `(batch_size,)`, the cross entropy loss is
        used for the single-label classification. In this case, labels should contain the indices that should be in
        `[0, ..., config.num_labels - 1]`. If the shape is `(batch_size, num_labels)`, the binary cross entropy
        loss is used for the multi-label classification. In this case, labels should only contain `[0, 1]`, where 0
        and 1 indicate false and true, respectively.

    Returns:

    Examples:

    ```python
    >>> from transformers import AutoTokenizer, LukeForEntityPairClassification

    >>> tokenizer = AutoTokenizer.from_pretrained("studio-ousia/luke-large-finetuned-tacred")
    >>> model = LukeForEntityPairClassification.from_pretrained("studio-ousia/luke-large-finetuned-tacred")

    >>> text = "Beyoncé lives in Los Angeles."
    >>> entity_spans = [
    ...     (0, 7),
    ...     (17, 28),
    ... ]  # character-based entity spans corresponding to "Beyoncé" and "Los Angeles"
    >>> inputs = tokenizer(text, entity_spans=entity_spans, return_tensors="ms")
    >>> outputs = model(**inputs)
    >>> logits = outputs.logits
    >>> predicted_class_idx = logits.argmax(-1).item()
    >>> print("Predicted class:", model.config.id2label[predicted_class_idx])
    Predicted class: per:cities_of_residence
    ```"""
    return_dict = return_dict if return_dict is not None else self.config.use_return_dict

    outputs = self.luke(
        input_ids=input_ids,
        attention_mask=attention_mask,
        token_type_ids=token_type_ids,
        position_ids=position_ids,
        entity_ids=entity_ids,
        entity_attention_mask=entity_attention_mask,
        entity_token_type_ids=entity_token_type_ids,
        entity_position_ids=entity_position_ids,
        head_mask=head_mask,
        inputs_embeds=inputs_embeds,
        output_attentions=output_attentions,
        output_hidden_states=output_hidden_states,
        return_dict=True,
    )

    feature_vector = ops.cat(
        [outputs.entity_last_hidden_state[:, 0, :], outputs.entity_last_hidden_state[:, 1, :]], dim=1
    )
    feature_vector = self.dropout(feature_vector)
    logits = self.classifier(feature_vector)

    loss = None
    if labels is not None:
        # When the number of dimension of `labels` is 1, cross entropy is used as the loss function. The binary
        # cross entropy is used otherwise.
        if labels.ndim == 1:
            loss = nn.functional.cross_entropy(logits, labels)
        else:
            loss = nn.functional.binary_cross_entropy_with_logits(logits.view(-1), labels.view(-1).type_as(logits))

    if not return_dict:
        return tuple(
            v
            for v in [loss, logits, outputs.hidden_states, outputs.entity_hidden_states, outputs.attentions]
            if v is not None
        )

    return EntityPairClassificationOutput(
        loss=loss,
        logits=logits,
        hidden_states=outputs.hidden_states,
        entity_hidden_states=outputs.entity_hidden_states,
        attentions=outputs.attentions,
    )

mindnlp.transformers.models.luke.modeling_luke.LukeForEntitySpanClassification

Bases: LukePreTrainedModel

Source code in mindnlp\transformers\models\luke\modeling_luke.py 
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
class LukeForEntitySpanClassification(LukePreTrainedModel):
    def __init__(self, config):
        super().__init__(config)

        self.luke = LukeModel(config)

        self.num_labels = config.num_labels
        self.dropout = nn.Dropout(config.hidden_dropout_prob)
        self.classifier = nn.Linear(config.hidden_size * 3, config.num_labels)

        # Initialize weights and apply final processing
        self.post_init()

    def forward(
        self,
        input_ids: Optional[mindspore.Tensor] = None,
        attention_mask: Optional[mindspore.Tensor] = None,
        token_type_ids: Optional[mindspore.Tensor] = None,
        position_ids: Optional[mindspore.Tensor] = None,
        entity_ids: Optional[mindspore.Tensor] = None,
        entity_attention_mask: Optional[mindspore.Tensor] = None,
        entity_token_type_ids: Optional[mindspore.Tensor] = None,
        entity_position_ids: Optional[mindspore.Tensor] = None,
        entity_start_positions: Optional[mindspore.Tensor] = None,
        entity_end_positions: Optional[mindspore.Tensor] = None,
        head_mask: Optional[mindspore.Tensor] = None,
        inputs_embeds: Optional[mindspore.Tensor] = None,
        labels: Optional[mindspore.Tensor] = None,
        output_attentions: Optional[bool] = None,
        output_hidden_states: Optional[bool] = None,
        return_dict: Optional[bool] = None,
    ) -> Union[Tuple, EntitySpanClassificationOutput]:
        r"""
        entity_start_positions (`mindspore.Tensor`):
            The start positions of entities in the word token sequence.

        entity_end_positions (`mindspore.Tensor`):
            The end positions of entities in the word token sequence.

        labels (`mindspore.Tensor` of shape `(batch_size, entity_length)` or `(batch_size, entity_length, num_labels)`, *optional*):
            Labels for computing the classification loss. If the shape is `(batch_size, entity_length)`, the cross
            entropy loss is used for the single-label classification. In this case, labels should contain the indices
            that should be in `[0, ..., config.num_labels - 1]`. If the shape is `(batch_size, entity_length,
            num_labels)`, the binary cross entropy loss is used for the multi-label classification. In this case,
            labels should only contain `[0, 1]`, where 0 and 1 indicate false and true, respectively.

        Returns:

        Examples:

        ```python
        >>> from transformers import AutoTokenizer, LukeForEntitySpanClassification

        >>> tokenizer = AutoTokenizer.from_pretrained("studio-ousia/luke-large-finetuned-conll-2003")
        >>> model = LukeForEntitySpanClassification.from_pretrained("studio-ousia/luke-large-finetuned-conll-2003")

        >>> text = "Beyoncé lives in Los Angeles"
        # List all possible entity spans in the text

        >>> word_start_positions = [0, 8, 14, 17, 21]  # character-based start positions of word tokens
        >>> word_end_positions = [7, 13, 16, 20, 28]  # character-based end positions of word tokens
        >>> entity_spans = []
        >>> for i, start_pos in enumerate(word_start_positions):
        ...     for end_pos in word_end_positions[i:]:
        ...         entity_spans.append((start_pos, end_pos))

        >>> inputs = tokenizer(text, entity_spans=entity_spans, return_tensors="ms")
        >>> outputs = model(**inputs)
        >>> logits = outputs.logits
        >>> predicted_class_indices = logits.argmax(-1).squeeze().tolist()
        >>> for span, predicted_class_idx in zip(entity_spans, predicted_class_indices):
        ...     if predicted_class_idx != 0:
        ...         print(text[span[0] : span[1]], model.config.id2label[predicted_class_idx])
        Beyoncé PER
        Los Angeles LOC
        ```"""
        return_dict = return_dict if return_dict is not None else self.config.use_return_dict

        outputs = self.luke(
            input_ids=input_ids,
            attention_mask=attention_mask,
            token_type_ids=token_type_ids,
            position_ids=position_ids,
            entity_ids=entity_ids,
            entity_attention_mask=entity_attention_mask,
            entity_token_type_ids=entity_token_type_ids,
            entity_position_ids=entity_position_ids,
            head_mask=head_mask,
            inputs_embeds=inputs_embeds,
            output_attentions=output_attentions,
            output_hidden_states=output_hidden_states,
            return_dict=True,
        )
        hidden_size = outputs.last_hidden_state.shape[-1]

        entity_start_positions = entity_start_positions.unsqueeze(-1).broadcast_to((-1, -1, hidden_size))
        start_states = ops.gather(outputs.last_hidden_state, -2, entity_start_positions)

        entity_end_positions = entity_end_positions.unsqueeze(-1).broadcast_to((-1, -1, hidden_size))
        end_states = ops.gather(outputs.last_hidden_state, -2, entity_end_positions)

        feature_vector = ops.cat([start_states, end_states, outputs.entity_last_hidden_state], dim=2)

        feature_vector = self.dropout(feature_vector)
        logits = self.classifier(feature_vector)

        loss = None
        if labels is not None:
            # When the number of dimension of `labels` is 2, cross entropy is used as the loss function. The binary
            # cross entropy is used otherwise.
            if labels.ndim == 2:
                loss = nn.functional.cross_entropy(logits.view(-1, self.num_labels), labels.view(-1))
            else:
                loss = nn.functional.binary_cross_entropy_with_logits(logits.view(-1), labels.view(-1).type_as(logits))

        if not return_dict:
            return tuple(
                v
                for v in [loss, logits, outputs.hidden_states, outputs.entity_hidden_states, outputs.attentions]
                if v is not None
            )

        return EntitySpanClassificationOutput(
            loss=loss,
            logits=logits,
            hidden_states=outputs.hidden_states,
            entity_hidden_states=outputs.entity_hidden_states,
            attentions=outputs.attentions,
        )

mindnlp.transformers.models.luke.modeling_luke.LukeForEntitySpanClassification.forward(input_ids=None, attention_mask=None, token_type_ids=None, position_ids=None, entity_ids=None, entity_attention_mask=None, entity_token_type_ids=None, entity_position_ids=None, entity_start_positions=None, entity_end_positions=None, head_mask=None, inputs_embeds=None, labels=None, output_attentions=None, output_hidden_states=None, return_dict=None)

entity_start_positions (mindspore.Tensor): The start positions of entities in the word token sequence.

entity_end_positions (mindspore.Tensor): The end positions of entities in the word token sequence.

labels (mindspore.Tensor of shape (batch_size, entity_length) or (batch_size, entity_length, num_labels), optional): Labels for computing the classification loss. If the shape is (batch_size, entity_length), the cross entropy loss is used for the single-label classification. In this case, labels should contain the indices that should be in [0, ..., config.num_labels - 1]. If the shape is (batch_size, entity_length, num_labels), the binary cross entropy loss is used for the multi-label classification. In this case, labels should only contain [0, 1], where 0 and 1 indicate false and true, respectively.

Returns:

Examples:

>>> from transformers import AutoTokenizer, LukeForEntitySpanClassification

>>> tokenizer = AutoTokenizer.from_pretrained("studio-ousia/luke-large-finetuned-conll-2003")
>>> model = LukeForEntitySpanClassification.from_pretrained("studio-ousia/luke-large-finetuned-conll-2003")

>>> text = "Beyoncé lives in Los Angeles"
# List all possible entity spans in the text

>>> word_start_positions = [0, 8, 14, 17, 21]  # character-based start positions of word tokens
>>> word_end_positions = [7, 13, 16, 20, 28]  # character-based end positions of word tokens
>>> entity_spans = []
>>> for i, start_pos in enumerate(word_start_positions):
...     for end_pos in word_end_positions[i:]:
...         entity_spans.append((start_pos, end_pos))

>>> inputs = tokenizer(text, entity_spans=entity_spans, return_tensors="ms")
>>> outputs = model(**inputs)
>>> logits = outputs.logits
>>> predicted_class_indices = logits.argmax(-1).squeeze().tolist()
>>> for span, predicted_class_idx in zip(entity_spans, predicted_class_indices):
...     if predicted_class_idx != 0:
...         print(text[span[0] : span[1]], model.config.id2label[predicted_class_idx])
Beyoncé PER
Los Angeles LOC
Source code in mindnlp\transformers\models\luke\modeling_luke.py 
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
def forward(
    self,
    input_ids: Optional[mindspore.Tensor] = None,
    attention_mask: Optional[mindspore.Tensor] = None,
    token_type_ids: Optional[mindspore.Tensor] = None,
    position_ids: Optional[mindspore.Tensor] = None,
    entity_ids: Optional[mindspore.Tensor] = None,
    entity_attention_mask: Optional[mindspore.Tensor] = None,
    entity_token_type_ids: Optional[mindspore.Tensor] = None,
    entity_position_ids: Optional[mindspore.Tensor] = None,
    entity_start_positions: Optional[mindspore.Tensor] = None,
    entity_end_positions: Optional[mindspore.Tensor] = None,
    head_mask: Optional[mindspore.Tensor] = None,
    inputs_embeds: Optional[mindspore.Tensor] = None,
    labels: Optional[mindspore.Tensor] = None,
    output_attentions: Optional[bool] = None,
    output_hidden_states: Optional[bool] = None,
    return_dict: Optional[bool] = None,
) -> Union[Tuple, EntitySpanClassificationOutput]:
    r"""
    entity_start_positions (`mindspore.Tensor`):
        The start positions of entities in the word token sequence.

    entity_end_positions (`mindspore.Tensor`):
        The end positions of entities in the word token sequence.

    labels (`mindspore.Tensor` of shape `(batch_size, entity_length)` or `(batch_size, entity_length, num_labels)`, *optional*):
        Labels for computing the classification loss. If the shape is `(batch_size, entity_length)`, the cross
        entropy loss is used for the single-label classification. In this case, labels should contain the indices
        that should be in `[0, ..., config.num_labels - 1]`. If the shape is `(batch_size, entity_length,
        num_labels)`, the binary cross entropy loss is used for the multi-label classification. In this case,
        labels should only contain `[0, 1]`, where 0 and 1 indicate false and true, respectively.

    Returns:

    Examples:

    ```python
    >>> from transformers import AutoTokenizer, LukeForEntitySpanClassification

    >>> tokenizer = AutoTokenizer.from_pretrained("studio-ousia/luke-large-finetuned-conll-2003")
    >>> model = LukeForEntitySpanClassification.from_pretrained("studio-ousia/luke-large-finetuned-conll-2003")

    >>> text = "Beyoncé lives in Los Angeles"
    # List all possible entity spans in the text

    >>> word_start_positions = [0, 8, 14, 17, 21]  # character-based start positions of word tokens
    >>> word_end_positions = [7, 13, 16, 20, 28]  # character-based end positions of word tokens
    >>> entity_spans = []
    >>> for i, start_pos in enumerate(word_start_positions):
    ...     for end_pos in word_end_positions[i:]:
    ...         entity_spans.append((start_pos, end_pos))

    >>> inputs = tokenizer(text, entity_spans=entity_spans, return_tensors="ms")
    >>> outputs = model(**inputs)
    >>> logits = outputs.logits
    >>> predicted_class_indices = logits.argmax(-1).squeeze().tolist()
    >>> for span, predicted_class_idx in zip(entity_spans, predicted_class_indices):
    ...     if predicted_class_idx != 0:
    ...         print(text[span[0] : span[1]], model.config.id2label[predicted_class_idx])
    Beyoncé PER
    Los Angeles LOC
    ```"""
    return_dict = return_dict if return_dict is not None else self.config.use_return_dict

    outputs = self.luke(
        input_ids=input_ids,
        attention_mask=attention_mask,
        token_type_ids=token_type_ids,
        position_ids=position_ids,
        entity_ids=entity_ids,
        entity_attention_mask=entity_attention_mask,
        entity_token_type_ids=entity_token_type_ids,
        entity_position_ids=entity_position_ids,
        head_mask=head_mask,
        inputs_embeds=inputs_embeds,
        output_attentions=output_attentions,
        output_hidden_states=output_hidden_states,
        return_dict=True,
    )
    hidden_size = outputs.last_hidden_state.shape[-1]

    entity_start_positions = entity_start_positions.unsqueeze(-1).broadcast_to((-1, -1, hidden_size))
    start_states = ops.gather(outputs.last_hidden_state, -2, entity_start_positions)

    entity_end_positions = entity_end_positions.unsqueeze(-1).broadcast_to((-1, -1, hidden_size))
    end_states = ops.gather(outputs.last_hidden_state, -2, entity_end_positions)

    feature_vector = ops.cat([start_states, end_states, outputs.entity_last_hidden_state], dim=2)

    feature_vector = self.dropout(feature_vector)
    logits = self.classifier(feature_vector)

    loss = None
    if labels is not None:
        # When the number of dimension of `labels` is 2, cross entropy is used as the loss function. The binary
        # cross entropy is used otherwise.
        if labels.ndim == 2:
            loss = nn.functional.cross_entropy(logits.view(-1, self.num_labels), labels.view(-1))
        else:
            loss = nn.functional.binary_cross_entropy_with_logits(logits.view(-1), labels.view(-1).type_as(logits))

    if not return_dict:
        return tuple(
            v
            for v in [loss, logits, outputs.hidden_states, outputs.entity_hidden_states, outputs.attentions]
            if v is not None
        )

    return EntitySpanClassificationOutput(
        loss=loss,
        logits=logits,
        hidden_states=outputs.hidden_states,
        entity_hidden_states=outputs.entity_hidden_states,
        attentions=outputs.attentions,
    )

mindnlp.transformers.models.luke.modeling_luke.LukeForMaskedLM

Bases: LukePreTrainedModel

Source code in mindnlp\transformers\models\luke\modeling_luke.py 
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
class LukeForMaskedLM(LukePreTrainedModel):
    _tied_weights_keys = ["lm_head.decoder.weight", "lm_head.decoder.bias", "entity_predictions.decoder.weight"]

    def __init__(self, config):
        super().__init__(config)

        self.luke = LukeModel(config)

        self.lm_head = LukeLMHead(config)
        self.entity_predictions = EntityPredictionHead(config)

        self.loss_fn = nn.CrossEntropyLoss()

        # Initialize weights and apply final processing
        self.post_init()

    def tie_weights(self):
        super().tie_weights()
        self._tie_or_clone_weights(self.entity_predictions.decoder, self.luke.entity_embeddings.entity_embeddings)

    def get_output_embeddings(self):
        return self.lm_head.decoder

    def set_output_embeddings(self, new_embeddings):
        self.lm_head.decoder = new_embeddings

    def forward(
        self,
        input_ids: Optional[mindspore.Tensor] = None,
        attention_mask: Optional[mindspore.Tensor] = None,
        token_type_ids: Optional[mindspore.Tensor] = None,
        position_ids: Optional[mindspore.Tensor] = None,
        entity_ids: Optional[mindspore.Tensor] = None,
        entity_attention_mask: Optional[mindspore.Tensor] = None,
        entity_token_type_ids: Optional[mindspore.Tensor] = None,
        entity_position_ids: Optional[mindspore.Tensor] = None,
        labels: Optional[mindspore.Tensor] = None,
        entity_labels: Optional[mindspore.Tensor] = None,
        head_mask: Optional[mindspore.Tensor] = None,
        inputs_embeds: Optional[mindspore.Tensor] = None,
        output_attentions: Optional[bool] = None,
        output_hidden_states: Optional[bool] = None,
        return_dict: Optional[bool] = None,
    ) -> Union[Tuple, LukeMaskedLMOutput]:
        r"""
        labels (`mindspore.Tensor` of shape `(batch_size, sequence_length)`, *optional*):
            Labels for computing the masked language modeling loss. Indices should be in `[-100, 0, ...,
            config.vocab_size]` (see `input_ids` docstring) Tokens with indices set to `-100` are ignored (masked), the
            loss is only computed for the tokens with labels in `[0, ..., config.vocab_size]`
        entity_labels (`mindspore.Tensor` of shape `(batch_size, entity_length)`, *optional*):
            Labels for computing the masked language modeling loss. Indices should be in `[-100, 0, ...,
            config.vocab_size]` (see `input_ids` docstring) Tokens with indices set to `-100` are ignored (masked), the
            loss is only computed for the tokens with labels in `[0, ..., config.vocab_size]`

        Returns:

        """

        return_dict = return_dict if return_dict is not None else self.config.use_return_dict

        outputs = self.luke(
            input_ids=input_ids,
            attention_mask=attention_mask,
            token_type_ids=token_type_ids,
            position_ids=position_ids,
            entity_ids=entity_ids,
            entity_attention_mask=entity_attention_mask,
            entity_token_type_ids=entity_token_type_ids,
            entity_position_ids=entity_position_ids,
            head_mask=head_mask,
            inputs_embeds=inputs_embeds,
            output_attentions=output_attentions,
            output_hidden_states=output_hidden_states,
            return_dict=True,
        )

        loss = None

        mlm_loss = None
        logits = self.lm_head(outputs.last_hidden_state)
        if labels is not None:
            mlm_loss = self.loss_fn(logits.view(-1, self.config.vocab_size), labels.view(-1))
            if loss is None:
                loss = mlm_loss

        mep_loss = None
        entity_logits = None
        if outputs.entity_last_hidden_state is not None:
            entity_logits = self.entity_predictions(outputs.entity_last_hidden_state)
            if entity_labels is not None:
                mep_loss = self.loss_fn(entity_logits.view(-1, self.config.entity_vocab_size), entity_labels.view(-1))
                if loss is None:
                    loss = mep_loss
                else:
                    loss = loss + mep_loss

        if not return_dict:
            return tuple(
                v
                for v in [
                    loss,
                    mlm_loss,
                    mep_loss,
                    logits,
                    entity_logits,
                    outputs.hidden_states,
                    outputs.entity_hidden_states,
                    outputs.attentions,
                ]
                if v is not None
            )

        return LukeMaskedLMOutput(
            loss=loss,
            mlm_loss=mlm_loss,
            mep_loss=mep_loss,
            logits=logits,
            entity_logits=entity_logits,
            hidden_states=outputs.hidden_states,
            entity_hidden_states=outputs.entity_hidden_states,
            attentions=outputs.attentions,
        )

mindnlp.transformers.models.luke.modeling_luke.LukeForMaskedLM.forward(input_ids=None, attention_mask=None, token_type_ids=None, position_ids=None, entity_ids=None, entity_attention_mask=None, entity_token_type_ids=None, entity_position_ids=None, labels=None, entity_labels=None, head_mask=None, inputs_embeds=None, output_attentions=None, output_hidden_states=None, return_dict=None)

labels (mindspore.Tensor of shape (batch_size, sequence_length), optional): Labels for computing the masked language modeling loss. Indices should be in [-100, 0, ..., config.vocab_size] (see input_ids docstring) Tokens with indices set to -100 are ignored (masked), the loss is only computed for the tokens with labels in [0, ..., config.vocab_size] entity_labels (mindspore.Tensor of shape (batch_size, entity_length), optional): Labels for computing the masked language modeling loss. Indices should be in [-100, 0, ..., config.vocab_size] (see input_ids docstring) Tokens with indices set to -100 are ignored (masked), the loss is only computed for the tokens with labels in [0, ..., config.vocab_size]

Returns:

Source code in mindnlp\transformers\models\luke\modeling_luke.py 
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
def forward(
    self,
    input_ids: Optional[mindspore.Tensor] = None,
    attention_mask: Optional[mindspore.Tensor] = None,
    token_type_ids: Optional[mindspore.Tensor] = None,
    position_ids: Optional[mindspore.Tensor] = None,
    entity_ids: Optional[mindspore.Tensor] = None,
    entity_attention_mask: Optional[mindspore.Tensor] = None,
    entity_token_type_ids: Optional[mindspore.Tensor] = None,
    entity_position_ids: Optional[mindspore.Tensor] = None,
    labels: Optional[mindspore.Tensor] = None,
    entity_labels: Optional[mindspore.Tensor] = None,
    head_mask: Optional[mindspore.Tensor] = None,
    inputs_embeds: Optional[mindspore.Tensor] = None,
    output_attentions: Optional[bool] = None,
    output_hidden_states: Optional[bool] = None,
    return_dict: Optional[bool] = None,
) -> Union[Tuple, LukeMaskedLMOutput]:
    r"""
    labels (`mindspore.Tensor` of shape `(batch_size, sequence_length)`, *optional*):
        Labels for computing the masked language modeling loss. Indices should be in `[-100, 0, ...,
        config.vocab_size]` (see `input_ids` docstring) Tokens with indices set to `-100` are ignored (masked), the
        loss is only computed for the tokens with labels in `[0, ..., config.vocab_size]`
    entity_labels (`mindspore.Tensor` of shape `(batch_size, entity_length)`, *optional*):
        Labels for computing the masked language modeling loss. Indices should be in `[-100, 0, ...,
        config.vocab_size]` (see `input_ids` docstring) Tokens with indices set to `-100` are ignored (masked), the
        loss is only computed for the tokens with labels in `[0, ..., config.vocab_size]`

    Returns:

    """

    return_dict = return_dict if return_dict is not None else self.config.use_return_dict

    outputs = self.luke(
        input_ids=input_ids,
        attention_mask=attention_mask,
        token_type_ids=token_type_ids,
        position_ids=position_ids,
        entity_ids=entity_ids,
        entity_attention_mask=entity_attention_mask,
        entity_token_type_ids=entity_token_type_ids,
        entity_position_ids=entity_position_ids,
        head_mask=head_mask,
        inputs_embeds=inputs_embeds,
        output_attentions=output_attentions,
        output_hidden_states=output_hidden_states,
        return_dict=True,
    )

    loss = None

    mlm_loss = None
    logits = self.lm_head(outputs.last_hidden_state)
    if labels is not None:
        mlm_loss = self.loss_fn(logits.view(-1, self.config.vocab_size), labels.view(-1))
        if loss is None:
            loss = mlm_loss

    mep_loss = None
    entity_logits = None
    if outputs.entity_last_hidden_state is not None:
        entity_logits = self.entity_predictions(outputs.entity_last_hidden_state)
        if entity_labels is not None:
            mep_loss = self.loss_fn(entity_logits.view(-1, self.config.entity_vocab_size), entity_labels.view(-1))
            if loss is None:
                loss = mep_loss
            else:
                loss = loss + mep_loss

    if not return_dict:
        return tuple(
            v
            for v in [
                loss,
                mlm_loss,
                mep_loss,
                logits,
                entity_logits,
                outputs.hidden_states,
                outputs.entity_hidden_states,
                outputs.attentions,
            ]
            if v is not None
        )

    return LukeMaskedLMOutput(
        loss=loss,
        mlm_loss=mlm_loss,
        mep_loss=mep_loss,
        logits=logits,
        entity_logits=entity_logits,
        hidden_states=outputs.hidden_states,
        entity_hidden_states=outputs.entity_hidden_states,
        attentions=outputs.attentions,
    )

mindnlp.transformers.models.luke.modeling_luke.LukeForMultipleChoice

Bases: LukePreTrainedModel

Source code in mindnlp\transformers\models\luke\modeling_luke.py 
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
class LukeForMultipleChoice(LukePreTrainedModel):
    def __init__(self, config):
        super().__init__(config)

        self.luke = LukeModel(config)
        self.dropout = nn.Dropout(
            config.classifier_dropout if config.classifier_dropout is not None else config.hidden_dropout_prob
        )
        self.classifier = nn.Linear(config.hidden_size, 1)

        # Initialize weights and apply final processing
        self.post_init()

    def forward(
        self,
        input_ids: Optional[mindspore.Tensor] = None,
        attention_mask: Optional[mindspore.Tensor] = None,
        token_type_ids: Optional[mindspore.Tensor] = None,
        position_ids: Optional[mindspore.Tensor] = None,
        entity_ids: Optional[mindspore.Tensor] = None,
        entity_attention_mask: Optional[mindspore.Tensor] = None,
        entity_token_type_ids: Optional[mindspore.Tensor] = None,
        entity_position_ids: Optional[mindspore.Tensor] = None,
        head_mask: Optional[mindspore.Tensor] = None,
        inputs_embeds: Optional[mindspore.Tensor] = None,
        labels: Optional[mindspore.Tensor] = None,
        output_attentions: Optional[bool] = None,
        output_hidden_states: Optional[bool] = None,
        return_dict: Optional[bool] = None,
    ) -> Union[Tuple, LukeMultipleChoiceModelOutput]:
        r"""
        labels (`mindspore.Tensor` of shape `(batch_size,)`, *optional*):
            Labels for computing the multiple choice classification loss. Indices should be in `[0, ...,
            num_choices-1]` where `num_choices` is the size of the second dimension of the input tensors. (See
            `input_ids` above)
        """
        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
        num_choices = input_ids.shape[1] if input_ids is not None else inputs_embeds.shape[1]

        input_ids = input_ids.view(-1, input_ids.shape[-1]) if input_ids is not None else None
        attention_mask = attention_mask.view(-1, attention_mask.shape[-1]) if attention_mask is not None else None
        token_type_ids = token_type_ids.view(-1, token_type_ids.shape[-1]) if token_type_ids is not None else None
        position_ids = position_ids.view(-1, position_ids.shape[-1]) if position_ids is not None else None
        inputs_embeds = (
            inputs_embeds.view(-1, inputs_embeds.shape[-2], inputs_embeds.shape[-1])
            if inputs_embeds is not None
            else None
        )

        entity_ids = entity_ids.view(-1, entity_ids.shape[-1]) if entity_ids is not None else None
        entity_attention_mask = (
            entity_attention_mask.view(-1, entity_attention_mask.shape[-1])
            if entity_attention_mask is not None
            else None
        )
        entity_token_type_ids = (
            entity_token_type_ids.view(-1, entity_token_type_ids.shape[-1])
            if entity_token_type_ids is not None
            else None
        )
        entity_position_ids = (
            entity_position_ids.view(-1, entity_position_ids.shape[-2], entity_position_ids.shape[-1])
            if entity_position_ids is not None
            else None
        )

        outputs = self.luke(
            input_ids=input_ids,
            attention_mask=attention_mask,
            token_type_ids=token_type_ids,
            position_ids=position_ids,
            entity_ids=entity_ids,
            entity_attention_mask=entity_attention_mask,
            entity_token_type_ids=entity_token_type_ids,
            entity_position_ids=entity_position_ids,
            head_mask=head_mask,
            inputs_embeds=inputs_embeds,
            output_attentions=output_attentions,
            output_hidden_states=output_hidden_states,
            return_dict=True,
        )

        pooled_output = outputs.pooler_output

        pooled_output = self.dropout(pooled_output)
        logits = self.classifier(pooled_output)
        reshaped_logits = logits.view(-1, num_choices)

        loss = None
        if labels is not None:
            loss_fct = CrossEntropyLoss()
            loss = loss_fct(reshaped_logits, labels)

        if not return_dict:
            return tuple(
                v
                for v in [
                    loss,
                    reshaped_logits,
                    outputs.hidden_states,
                    outputs.entity_hidden_states,
                    outputs.attentions,
                ]
                if v is not None
            )

        return LukeMultipleChoiceModelOutput(
            loss=loss,
            logits=reshaped_logits,
            hidden_states=outputs.hidden_states,
            entity_hidden_states=outputs.entity_hidden_states,
            attentions=outputs.attentions,
        )

mindnlp.transformers.models.luke.modeling_luke.LukeForMultipleChoice.forward(input_ids=None, attention_mask=None, token_type_ids=None, position_ids=None, entity_ids=None, entity_attention_mask=None, entity_token_type_ids=None, entity_position_ids=None, head_mask=None, inputs_embeds=None, labels=None, output_attentions=None, output_hidden_states=None, return_dict=None)

labels (mindspore.Tensor of shape (batch_size,), optional): Labels for computing the multiple choice classification loss. Indices should be in [0, ..., num_choices-1] where num_choices is the size of the second dimension of the input tensors. (See input_ids above)

Source code in mindnlp\transformers\models\luke\modeling_luke.py 
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
def forward(
    self,
    input_ids: Optional[mindspore.Tensor] = None,
    attention_mask: Optional[mindspore.Tensor] = None,
    token_type_ids: Optional[mindspore.Tensor] = None,
    position_ids: Optional[mindspore.Tensor] = None,
    entity_ids: Optional[mindspore.Tensor] = None,
    entity_attention_mask: Optional[mindspore.Tensor] = None,
    entity_token_type_ids: Optional[mindspore.Tensor] = None,
    entity_position_ids: Optional[mindspore.Tensor] = None,
    head_mask: Optional[mindspore.Tensor] = None,
    inputs_embeds: Optional[mindspore.Tensor] = None,
    labels: Optional[mindspore.Tensor] = None,
    output_attentions: Optional[bool] = None,
    output_hidden_states: Optional[bool] = None,
    return_dict: Optional[bool] = None,
) -> Union[Tuple, LukeMultipleChoiceModelOutput]:
    r"""
    labels (`mindspore.Tensor` of shape `(batch_size,)`, *optional*):
        Labels for computing the multiple choice classification loss. Indices should be in `[0, ...,
        num_choices-1]` where `num_choices` is the size of the second dimension of the input tensors. (See
        `input_ids` above)
    """
    return_dict = return_dict if return_dict is not None else self.config.use_return_dict
    num_choices = input_ids.shape[1] if input_ids is not None else inputs_embeds.shape[1]

    input_ids = input_ids.view(-1, input_ids.shape[-1]) if input_ids is not None else None
    attention_mask = attention_mask.view(-1, attention_mask.shape[-1]) if attention_mask is not None else None
    token_type_ids = token_type_ids.view(-1, token_type_ids.shape[-1]) if token_type_ids is not None else None
    position_ids = position_ids.view(-1, position_ids.shape[-1]) if position_ids is not None else None
    inputs_embeds = (
        inputs_embeds.view(-1, inputs_embeds.shape[-2], inputs_embeds.shape[-1])
        if inputs_embeds is not None
        else None
    )

    entity_ids = entity_ids.view(-1, entity_ids.shape[-1]) if entity_ids is not None else None
    entity_attention_mask = (
        entity_attention_mask.view(-1, entity_attention_mask.shape[-1])
        if entity_attention_mask is not None
        else None
    )
    entity_token_type_ids = (
        entity_token_type_ids.view(-1, entity_token_type_ids.shape[-1])
        if entity_token_type_ids is not None
        else None
    )
    entity_position_ids = (
        entity_position_ids.view(-1, entity_position_ids.shape[-2], entity_position_ids.shape[-1])
        if entity_position_ids is not None
        else None
    )

    outputs = self.luke(
        input_ids=input_ids,
        attention_mask=attention_mask,
        token_type_ids=token_type_ids,
        position_ids=position_ids,
        entity_ids=entity_ids,
        entity_attention_mask=entity_attention_mask,
        entity_token_type_ids=entity_token_type_ids,
        entity_position_ids=entity_position_ids,
        head_mask=head_mask,
        inputs_embeds=inputs_embeds,
        output_attentions=output_attentions,
        output_hidden_states=output_hidden_states,
        return_dict=True,
    )

    pooled_output = outputs.pooler_output

    pooled_output = self.dropout(pooled_output)
    logits = self.classifier(pooled_output)
    reshaped_logits = logits.view(-1, num_choices)

    loss = None
    if labels is not None:
        loss_fct = CrossEntropyLoss()
        loss = loss_fct(reshaped_logits, labels)

    if not return_dict:
        return tuple(
            v
            for v in [
                loss,
                reshaped_logits,
                outputs.hidden_states,
                outputs.entity_hidden_states,
                outputs.attentions,
            ]
            if v is not None
        )

    return LukeMultipleChoiceModelOutput(
        loss=loss,
        logits=reshaped_logits,
        hidden_states=outputs.hidden_states,
        entity_hidden_states=outputs.entity_hidden_states,
        attentions=outputs.attentions,
    )

mindnlp.transformers.models.luke.modeling_luke.LukeForQuestionAnswering

Bases: LukePreTrainedModel

Source code in mindnlp\transformers\models\luke\modeling_luke.py 
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
1845
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
1882
1883
1884
1885
1886
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
class LukeForQuestionAnswering(LukePreTrainedModel):
    def __init__(self, config):
        super().__init__(config)

        self.num_labels = config.num_labels

        self.luke = LukeModel(config, add_pooling_layer=False)
        self.qa_outputs = nn.Linear(config.hidden_size, config.num_labels)

        # Initialize weights and apply final processing
        self.post_init()

    def forward(
        self,
        input_ids: Optional[mindspore.Tensor] = None,
        attention_mask: Optional[mindspore.Tensor] = None,
        token_type_ids: Optional[mindspore.Tensor] = None,
        position_ids: Optional[mindspore.Tensor] = None,
        entity_ids: Optional[mindspore.Tensor] = None,
        entity_attention_mask: Optional[mindspore.Tensor] = None,
        entity_token_type_ids: Optional[mindspore.Tensor] = None,
        entity_position_ids: Optional[mindspore.Tensor] = None,
        head_mask: Optional[mindspore.Tensor] = None,
        inputs_embeds: Optional[mindspore.Tensor] = None,
        start_positions: Optional[mindspore.Tensor] = None,
        end_positions: Optional[mindspore.Tensor] = None,
        output_attentions: Optional[bool] = None,
        output_hidden_states: Optional[bool] = None,
        return_dict: Optional[bool] = None,
    ) -> Union[Tuple, LukeQuestionAnsweringModelOutput]:
        r"""
        start_positions (`mindspore.Tensor` of shape `(batch_size,)`, *optional*):
            Labels for position (index) of the start of the labelled span for computing the token classification loss.
            Positions are clamped to the length of the sequence (`sequence_length`). Position outside of the sequence
            are not taken into account for computing the loss.
        end_positions (`mindspore.Tensor` of shape `(batch_size,)`, *optional*):
            Labels for position (index) of the end of the labelled span for computing the token classification loss.
            Positions are clamped to the length of the sequence (`sequence_length`). Position outside of the sequence
            are not taken into account for computing the loss.
        """
        return_dict = return_dict if return_dict is not None else self.config.use_return_dict

        outputs = self.luke(
            input_ids=input_ids,
            attention_mask=attention_mask,
            token_type_ids=token_type_ids,
            position_ids=position_ids,
            entity_ids=entity_ids,
            entity_attention_mask=entity_attention_mask,
            entity_token_type_ids=entity_token_type_ids,
            entity_position_ids=entity_position_ids,
            head_mask=head_mask,
            inputs_embeds=inputs_embeds,
            output_attentions=output_attentions,
            output_hidden_states=output_hidden_states,
            return_dict=True,
        )

        sequence_output = outputs.last_hidden_state

        logits = self.qa_outputs(sequence_output)
        start_logits, end_logits = logits.split(1, axis=-1)
        start_logits = start_logits.squeeze(-1)
        end_logits = end_logits.squeeze(-1)

        total_loss = None
        if start_positions is not None and end_positions is not None:
            # If we are on multi-GPU, split add a dimension
            if len(start_positions.shape) > 1:
                start_positions = start_positions.squeeze(-1)
            if len(end_positions.shape) > 1:
                end_positions = end_positions.squeeze(-1)
            # sometimes the start/end positions are outside our model inputs, we ignore these terms
            ignored_index = start_logits.shape[1]
            start_positions = start_positions.clamp(0, ignored_index)
            end_positions = end_positions.clamp(0, ignored_index)

            loss_fct = CrossEntropyLoss(ignore_index=ignored_index)
            start_loss = loss_fct(start_logits, start_positions)
            end_loss = loss_fct(end_logits, end_positions)
            total_loss = (start_loss + end_loss) / 2

        if not return_dict:
            return tuple(
                v
                for v in [
                    total_loss,
                    start_logits,
                    end_logits,
                    outputs.hidden_states,
                    outputs.entity_hidden_states,
                    outputs.attentions,
                ]
                if v is not None
            )

        return LukeQuestionAnsweringModelOutput(
            loss=total_loss,
            start_logits=start_logits,
            end_logits=end_logits,
            hidden_states=outputs.hidden_states,
            entity_hidden_states=outputs.entity_hidden_states,
            attentions=outputs.attentions,
        )

mindnlp.transformers.models.luke.modeling_luke.LukeForQuestionAnswering.forward(input_ids=None, attention_mask=None, token_type_ids=None, position_ids=None, entity_ids=None, entity_attention_mask=None, entity_token_type_ids=None, entity_position_ids=None, head_mask=None, inputs_embeds=None, start_positions=None, end_positions=None, output_attentions=None, output_hidden_states=None, return_dict=None)

start_positions (mindspore.Tensor of shape (batch_size,), optional): Labels for position (index) of the start of the labelled span for computing the token classification loss. Positions are clamped to the length of the sequence (sequence_length). Position outside of the sequence are not taken into account for computing the loss. end_positions (mindspore.Tensor of shape (batch_size,), optional): Labels for position (index) of the end of the labelled span for computing the token classification loss. Positions are clamped to the length of the sequence (sequence_length). Position outside of the sequence are not taken into account for computing the loss.

Source code in mindnlp\transformers\models\luke\modeling_luke.py 
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
1845
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
1882
1883
1884
1885
1886
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
def forward(
    self,
    input_ids: Optional[mindspore.Tensor] = None,
    attention_mask: Optional[mindspore.Tensor] = None,
    token_type_ids: Optional[mindspore.Tensor] = None,
    position_ids: Optional[mindspore.Tensor] = None,
    entity_ids: Optional[mindspore.Tensor] = None,
    entity_attention_mask: Optional[mindspore.Tensor] = None,
    entity_token_type_ids: Optional[mindspore.Tensor] = None,
    entity_position_ids: Optional[mindspore.Tensor] = None,
    head_mask: Optional[mindspore.Tensor] = None,
    inputs_embeds: Optional[mindspore.Tensor] = None,
    start_positions: Optional[mindspore.Tensor] = None,
    end_positions: Optional[mindspore.Tensor] = None,
    output_attentions: Optional[bool] = None,
    output_hidden_states: Optional[bool] = None,
    return_dict: Optional[bool] = None,
) -> Union[Tuple, LukeQuestionAnsweringModelOutput]:
    r"""
    start_positions (`mindspore.Tensor` of shape `(batch_size,)`, *optional*):
        Labels for position (index) of the start of the labelled span for computing the token classification loss.
        Positions are clamped to the length of the sequence (`sequence_length`). Position outside of the sequence
        are not taken into account for computing the loss.
    end_positions (`mindspore.Tensor` of shape `(batch_size,)`, *optional*):
        Labels for position (index) of the end of the labelled span for computing the token classification loss.
        Positions are clamped to the length of the sequence (`sequence_length`). Position outside of the sequence
        are not taken into account for computing the loss.
    """
    return_dict = return_dict if return_dict is not None else self.config.use_return_dict

    outputs = self.luke(
        input_ids=input_ids,
        attention_mask=attention_mask,
        token_type_ids=token_type_ids,
        position_ids=position_ids,
        entity_ids=entity_ids,
        entity_attention_mask=entity_attention_mask,
        entity_token_type_ids=entity_token_type_ids,
        entity_position_ids=entity_position_ids,
        head_mask=head_mask,
        inputs_embeds=inputs_embeds,
        output_attentions=output_attentions,
        output_hidden_states=output_hidden_states,
        return_dict=True,
    )

    sequence_output = outputs.last_hidden_state

    logits = self.qa_outputs(sequence_output)
    start_logits, end_logits = logits.split(1, axis=-1)
    start_logits = start_logits.squeeze(-1)
    end_logits = end_logits.squeeze(-1)

    total_loss = None
    if start_positions is not None and end_positions is not None:
        # If we are on multi-GPU, split add a dimension
        if len(start_positions.shape) > 1:
            start_positions = start_positions.squeeze(-1)
        if len(end_positions.shape) > 1:
            end_positions = end_positions.squeeze(-1)
        # sometimes the start/end positions are outside our model inputs, we ignore these terms
        ignored_index = start_logits.shape[1]
        start_positions = start_positions.clamp(0, ignored_index)
        end_positions = end_positions.clamp(0, ignored_index)

        loss_fct = CrossEntropyLoss(ignore_index=ignored_index)
        start_loss = loss_fct(start_logits, start_positions)
        end_loss = loss_fct(end_logits, end_positions)
        total_loss = (start_loss + end_loss) / 2

    if not return_dict:
        return tuple(
            v
            for v in [
                total_loss,
                start_logits,
                end_logits,
                outputs.hidden_states,
                outputs.entity_hidden_states,
                outputs.attentions,
            ]
            if v is not None
        )

    return LukeQuestionAnsweringModelOutput(
        loss=total_loss,
        start_logits=start_logits,
        end_logits=end_logits,
        hidden_states=outputs.hidden_states,
        entity_hidden_states=outputs.entity_hidden_states,
        attentions=outputs.attentions,
    )

mindnlp.transformers.models.luke.modeling_luke.LukeForSequenceClassification

Bases: LukePreTrainedModel

Source code in mindnlp\transformers\models\luke\modeling_luke.py 
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
class LukeForSequenceClassification(LukePreTrainedModel):
    def __init__(self, config):
        super().__init__(config)
        self.num_labels = config.num_labels
        self.luke = LukeModel(config)
        self.dropout = nn.Dropout(
            config.classifier_dropout if config.classifier_dropout is not None else config.hidden_dropout_prob
        )
        self.classifier = nn.Linear(config.hidden_size, config.num_labels)

        # Initialize weights and apply final processing
        self.post_init()

    def forward(
        self,
        input_ids: Optional[mindspore.Tensor] = None,
        attention_mask: Optional[mindspore.Tensor] = None,
        token_type_ids: Optional[mindspore.Tensor] = None,
        position_ids: Optional[mindspore.Tensor] = None,
        entity_ids: Optional[mindspore.Tensor] = None,
        entity_attention_mask: Optional[mindspore.Tensor] = None,
        entity_token_type_ids: Optional[mindspore.Tensor] = None,
        entity_position_ids: Optional[mindspore.Tensor] = None,
        head_mask: Optional[mindspore.Tensor] = None,
        inputs_embeds: Optional[mindspore.Tensor] = None,
        labels: Optional[mindspore.Tensor] = None,
        output_attentions: Optional[bool] = None,
        output_hidden_states: Optional[bool] = None,
        return_dict: Optional[bool] = None,
    ) -> Union[Tuple, LukeSequenceClassifierOutput]:
        r"""
        labels (`mindspore.Tensor` of shape `(batch_size,)`, *optional*):
            Labels for computing the sequence classification/regression loss. Indices should be in `[0, ...,
            config.num_labels - 1]`. If `config.num_labels == 1` a regression loss is computed (Mean-Square loss), If
            `config.num_labels > 1` a classification loss is computed (Cross-Entropy).
        """
        return_dict = return_dict if return_dict is not None else self.config.use_return_dict

        outputs = self.luke(
            input_ids=input_ids,
            attention_mask=attention_mask,
            token_type_ids=token_type_ids,
            position_ids=position_ids,
            entity_ids=entity_ids,
            entity_attention_mask=entity_attention_mask,
            entity_token_type_ids=entity_token_type_ids,
            entity_position_ids=entity_position_ids,
            head_mask=head_mask,
            inputs_embeds=inputs_embeds,
            output_attentions=output_attentions,
            output_hidden_states=output_hidden_states,
            return_dict=True,
        )

        pooled_output = outputs.pooler_output

        pooled_output = self.dropout(pooled_output)
        logits = self.classifier(pooled_output)

        loss = None
        if labels is not None:
            if self.config.problem_type is None:
                if self.num_labels == 1:
                    self.config.problem_type = "regression"
                elif self.num_labels > 1 and labels.dtype in (mindspore.int64, mindspore.int32):
                    self.config.problem_type = "single_label_classification"
                else:
                    self.config.problem_type = "multi_label_classification"

            if self.config.problem_type == "regression":
                loss_fct = MSELoss()
                if self.num_labels == 1:
                    loss = loss_fct(logits.squeeze(), labels.squeeze())
                else:
                    loss = loss_fct(logits, labels)
            elif self.config.problem_type == "single_label_classification":
                loss_fct = CrossEntropyLoss()
                loss = loss_fct(logits.view(-1, self.num_labels), labels.view(-1))
            elif self.config.problem_type == "multi_label_classification":
                loss_fct = BCEWithLogitsLoss()
                loss = loss_fct(logits, labels)

        if not return_dict:
            return tuple(
                v
                for v in [loss, logits, outputs.hidden_states, outputs.entity_hidden_states, outputs.attentions]
                if v is not None
            )

        return LukeSequenceClassifierOutput(
            loss=loss,
            logits=logits,
            hidden_states=outputs.hidden_states,
            entity_hidden_states=outputs.entity_hidden_states,
            attentions=outputs.attentions,
        )

mindnlp.transformers.models.luke.modeling_luke.LukeForSequenceClassification.forward(input_ids=None, attention_mask=None, token_type_ids=None, position_ids=None, entity_ids=None, entity_attention_mask=None, entity_token_type_ids=None, entity_position_ids=None, head_mask=None, inputs_embeds=None, labels=None, output_attentions=None, output_hidden_states=None, return_dict=None)

labels (mindspore.Tensor of shape (batch_size,), optional): Labels for computing the sequence classification/regression loss. Indices should be in [0, ..., config.num_labels - 1]. If config.num_labels == 1 a regression loss is computed (Mean-Square loss), If config.num_labels > 1 a classification loss is computed (Cross-Entropy).

Source code in mindnlp\transformers\models\luke\modeling_luke.py 
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
def forward(
    self,
    input_ids: Optional[mindspore.Tensor] = None,
    attention_mask: Optional[mindspore.Tensor] = None,
    token_type_ids: Optional[mindspore.Tensor] = None,
    position_ids: Optional[mindspore.Tensor] = None,
    entity_ids: Optional[mindspore.Tensor] = None,
    entity_attention_mask: Optional[mindspore.Tensor] = None,
    entity_token_type_ids: Optional[mindspore.Tensor] = None,
    entity_position_ids: Optional[mindspore.Tensor] = None,
    head_mask: Optional[mindspore.Tensor] = None,
    inputs_embeds: Optional[mindspore.Tensor] = None,
    labels: Optional[mindspore.Tensor] = None,
    output_attentions: Optional[bool] = None,
    output_hidden_states: Optional[bool] = None,
    return_dict: Optional[bool] = None,
) -> Union[Tuple, LukeSequenceClassifierOutput]:
    r"""
    labels (`mindspore.Tensor` of shape `(batch_size,)`, *optional*):
        Labels for computing the sequence classification/regression loss. Indices should be in `[0, ...,
        config.num_labels - 1]`. If `config.num_labels == 1` a regression loss is computed (Mean-Square loss), If
        `config.num_labels > 1` a classification loss is computed (Cross-Entropy).
    """
    return_dict = return_dict if return_dict is not None else self.config.use_return_dict

    outputs = self.luke(
        input_ids=input_ids,
        attention_mask=attention_mask,
        token_type_ids=token_type_ids,
        position_ids=position_ids,
        entity_ids=entity_ids,
        entity_attention_mask=entity_attention_mask,
        entity_token_type_ids=entity_token_type_ids,
        entity_position_ids=entity_position_ids,
        head_mask=head_mask,
        inputs_embeds=inputs_embeds,
        output_attentions=output_attentions,
        output_hidden_states=output_hidden_states,
        return_dict=True,
    )

    pooled_output = outputs.pooler_output

    pooled_output = self.dropout(pooled_output)
    logits = self.classifier(pooled_output)

    loss = None
    if labels is not None:
        if self.config.problem_type is None:
            if self.num_labels == 1:
                self.config.problem_type = "regression"
            elif self.num_labels > 1 and labels.dtype in (mindspore.int64, mindspore.int32):
                self.config.problem_type = "single_label_classification"
            else:
                self.config.problem_type = "multi_label_classification"

        if self.config.problem_type == "regression":
            loss_fct = MSELoss()
            if self.num_labels == 1:
                loss = loss_fct(logits.squeeze(), labels.squeeze())
            else:
                loss = loss_fct(logits, labels)
        elif self.config.problem_type == "single_label_classification":
            loss_fct = CrossEntropyLoss()
            loss = loss_fct(logits.view(-1, self.num_labels), labels.view(-1))
        elif self.config.problem_type == "multi_label_classification":
            loss_fct = BCEWithLogitsLoss()
            loss = loss_fct(logits, labels)

    if not return_dict:
        return tuple(
            v
            for v in [loss, logits, outputs.hidden_states, outputs.entity_hidden_states, outputs.attentions]
            if v is not None
        )

    return LukeSequenceClassifierOutput(
        loss=loss,
        logits=logits,
        hidden_states=outputs.hidden_states,
        entity_hidden_states=outputs.entity_hidden_states,
        attentions=outputs.attentions,
    )

mindnlp.transformers.models.luke.modeling_luke.LukeForTokenClassification

Bases: LukePreTrainedModel

Source code in mindnlp\transformers\models\luke\modeling_luke.py 
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
class LukeForTokenClassification(LukePreTrainedModel):
    def __init__(self, config):
        super().__init__(config)
        self.num_labels = config.num_labels

        self.luke = LukeModel(config, add_pooling_layer=False)
        self.dropout = nn.Dropout(
            config.classifier_dropout if config.classifier_dropout is not None else config.hidden_dropout_prob
        )
        self.classifier = nn.Linear(config.hidden_size, config.num_labels)

        # Initialize weights and apply final processing
        self.post_init()

    def forward(
        self,
        input_ids: Optional[mindspore.Tensor] = None,
        attention_mask: Optional[mindspore.Tensor] = None,
        token_type_ids: Optional[mindspore.Tensor] = None,
        position_ids: Optional[mindspore.Tensor] = None,
        entity_ids: Optional[mindspore.Tensor] = None,
        entity_attention_mask: Optional[mindspore.Tensor] = None,
        entity_token_type_ids: Optional[mindspore.Tensor] = None,
        entity_position_ids: Optional[mindspore.Tensor] = None,
        head_mask: Optional[mindspore.Tensor] = None,
        inputs_embeds: Optional[mindspore.Tensor] = None,
        labels: Optional[mindspore.Tensor] = None,
        output_attentions: Optional[bool] = None,
        output_hidden_states: Optional[bool] = None,
        return_dict: Optional[bool] = None,
    ) -> Union[Tuple, LukeTokenClassifierOutput]:
        r"""
        labels (`mindspore.Tensor` of shape `(batch_size,)`, *optional*):
            Labels for computing the multiple choice classification loss. Indices should be in `[0, ...,
            num_choices-1]` where `num_choices` is the size of the second dimension of the input tensors. (See
            `input_ids` above)
        """
        return_dict = return_dict if return_dict is not None else self.config.use_return_dict

        outputs = self.luke(
            input_ids=input_ids,
            attention_mask=attention_mask,
            token_type_ids=token_type_ids,
            position_ids=position_ids,
            entity_ids=entity_ids,
            entity_attention_mask=entity_attention_mask,
            entity_token_type_ids=entity_token_type_ids,
            entity_position_ids=entity_position_ids,
            head_mask=head_mask,
            inputs_embeds=inputs_embeds,
            output_attentions=output_attentions,
            output_hidden_states=output_hidden_states,
            return_dict=True,
        )

        sequence_output = outputs.last_hidden_state

        sequence_output = self.dropout(sequence_output)
        logits = self.classifier(sequence_output)

        loss = None
        if labels is not None:
            loss_fct = CrossEntropyLoss()
            loss = loss_fct(logits.view(-1, self.num_labels), labels.view(-1))

        if not return_dict:
            return tuple(
                v
                for v in [loss, logits, outputs.hidden_states, outputs.entity_hidden_states, outputs.attentions]
                if v is not None
            )

        return LukeTokenClassifierOutput(
            loss=loss,
            logits=logits,
            hidden_states=outputs.hidden_states,
            entity_hidden_states=outputs.entity_hidden_states,
            attentions=outputs.attentions,
        )

mindnlp.transformers.models.luke.modeling_luke.LukeForTokenClassification.forward(input_ids=None, attention_mask=None, token_type_ids=None, position_ids=None, entity_ids=None, entity_attention_mask=None, entity_token_type_ids=None, entity_position_ids=None, head_mask=None, inputs_embeds=None, labels=None, output_attentions=None, output_hidden_states=None, return_dict=None)

labels (mindspore.Tensor of shape (batch_size,), optional): Labels for computing the multiple choice classification loss. Indices should be in [0, ..., num_choices-1] where num_choices is the size of the second dimension of the input tensors. (See input_ids above)

Source code in mindnlp\transformers\models\luke\modeling_luke.py 
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
def forward(
    self,
    input_ids: Optional[mindspore.Tensor] = None,
    attention_mask: Optional[mindspore.Tensor] = None,
    token_type_ids: Optional[mindspore.Tensor] = None,
    position_ids: Optional[mindspore.Tensor] = None,
    entity_ids: Optional[mindspore.Tensor] = None,
    entity_attention_mask: Optional[mindspore.Tensor] = None,
    entity_token_type_ids: Optional[mindspore.Tensor] = None,
    entity_position_ids: Optional[mindspore.Tensor] = None,
    head_mask: Optional[mindspore.Tensor] = None,
    inputs_embeds: Optional[mindspore.Tensor] = None,
    labels: Optional[mindspore.Tensor] = None,
    output_attentions: Optional[bool] = None,
    output_hidden_states: Optional[bool] = None,
    return_dict: Optional[bool] = None,
) -> Union[Tuple, LukeTokenClassifierOutput]:
    r"""
    labels (`mindspore.Tensor` of shape `(batch_size,)`, *optional*):
        Labels for computing the multiple choice classification loss. Indices should be in `[0, ...,
        num_choices-1]` where `num_choices` is the size of the second dimension of the input tensors. (See
        `input_ids` above)
    """
    return_dict = return_dict if return_dict is not None else self.config.use_return_dict

    outputs = self.luke(
        input_ids=input_ids,
        attention_mask=attention_mask,
        token_type_ids=token_type_ids,
        position_ids=position_ids,
        entity_ids=entity_ids,
        entity_attention_mask=entity_attention_mask,
        entity_token_type_ids=entity_token_type_ids,
        entity_position_ids=entity_position_ids,
        head_mask=head_mask,
        inputs_embeds=inputs_embeds,
        output_attentions=output_attentions,
        output_hidden_states=output_hidden_states,
        return_dict=True,
    )

    sequence_output = outputs.last_hidden_state

    sequence_output = self.dropout(sequence_output)
    logits = self.classifier(sequence_output)

    loss = None
    if labels is not None:
        loss_fct = CrossEntropyLoss()
        loss = loss_fct(logits.view(-1, self.num_labels), labels.view(-1))

    if not return_dict:
        return tuple(
            v
            for v in [loss, logits, outputs.hidden_states, outputs.entity_hidden_states, outputs.attentions]
            if v is not None
        )

    return LukeTokenClassifierOutput(
        loss=loss,
        logits=logits,
        hidden_states=outputs.hidden_states,
        entity_hidden_states=outputs.entity_hidden_states,
        attentions=outputs.attentions,
    )

mindnlp.transformers.models.luke.modeling_luke.LukeLMHead

Bases: Module

Roberta Head for masked language modeling.

Source code in mindnlp\transformers\models\luke\modeling_luke.py 
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
class LukeLMHead(nn.Module):
    """Roberta Head for masked language modeling."""

    def __init__(self, config):
        super().__init__()
        self.dense = nn.Linear(config.hidden_size, config.hidden_size)
        self.layer_norm = nn.LayerNorm(config.hidden_size, eps=config.layer_norm_eps)

        self.decoder = nn.Linear(config.hidden_size, config.vocab_size)
        self.bias = nn.Parameter(ops.zeros(config.vocab_size))
        self.decoder.bias = self.bias

    def forward(self, features, **kwargs):
        x = self.dense(features)
        x = gelu(x)
        x = self.layer_norm(x)

        # project back to size of vocabulary with bias
        x = self.decoder(x)

        return x

    def _tie_weights(self):
        # To tie those two weights if they get disconnected (on TPU or when the bias is resized)
        # For accelerate compatibility and to not break backward compatibility
        self.bias = self.decoder.bias

mindnlp.transformers.models.luke.modeling_luke.LukeMaskedLMOutput dataclass

Bases: ModelOutput

Base class for model's outputs, with potential hidden states and attentions.

PARAMETER DESCRIPTION
loss

The sum of masked language modeling (MLM) loss and entity prediction loss.

TYPE: `mindspore.Tensor` of shape `(1,)`, *optional*, returned when `labels` is provided DEFAULT: None

mlm_loss

Masked language modeling (MLM) loss.

TYPE: `mindspore.Tensor` of shape `(1,)`, *optional*, returned when `labels` is provided DEFAULT: None

mep_loss

Masked entity prediction (MEP) loss.

TYPE: `mindspore.Tensor` of shape `(1,)`, *optional*, returned when `labels` is provided DEFAULT: None

logits

Prediction scores of the language modeling head (scores for each vocabulary token before SoftMax).

TYPE: `mindspore.Tensor` of shape `(batch_size, sequence_length, config.vocab_size)` DEFAULT: None

entity_logits

Prediction scores of the entity prediction head (scores for each entity vocabulary token before SoftMax).

TYPE: `mindspore.Tensor` of shape `(batch_size, sequence_length, config.vocab_size)` DEFAULT: None

hidden_states

Tuple of mindspore.Tensor (one for the output of the embeddings + one for the output of each layer) of shape (batch_size, sequence_length, hidden_size).

Hidden-states of the model at the output of each layer plus the initial embedding outputs.

TYPE: `tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True` DEFAULT: None

entity_hidden_states

Tuple of mindspore.Tensor (one for the output of the embeddings + one for the output of each layer) of shape (batch_size, entity_length, hidden_size). Entity hidden-states of the model at the output of each layer plus the initial entity embedding outputs.

TYPE: `tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True` DEFAULT: None

attentions

Tuple of mindspore.Tensor (one for each layer) of shape (batch_size, num_heads, sequence_length, sequence_length).

Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads.

TYPE: `tuple(mindspore.Tensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True` DEFAULT: None

Source code in mindnlp\transformers\models\luke\modeling_luke.py 
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
@dataclass
class LukeMaskedLMOutput(ModelOutput):
    """
    Base class for model's outputs, with potential hidden states and attentions.

    Args:
        loss (`mindspore.Tensor` of shape `(1,)`, *optional*, returned when `labels` is provided):
            The sum of masked language modeling (MLM) loss and entity prediction loss.
        mlm_loss (`mindspore.Tensor` of shape `(1,)`, *optional*, returned when `labels` is provided):
            Masked language modeling (MLM) loss.
        mep_loss (`mindspore.Tensor` of shape `(1,)`, *optional*, returned when `labels` is provided):
            Masked entity prediction (MEP) loss.
        logits (`mindspore.Tensor` of shape `(batch_size, sequence_length, config.vocab_size)`):
            Prediction scores of the language modeling head (scores for each vocabulary token before SoftMax).
        entity_logits (`mindspore.Tensor` of shape `(batch_size, sequence_length, config.vocab_size)`):
            Prediction scores of the entity prediction head (scores for each entity vocabulary token before SoftMax).
        hidden_states (`tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`):
            Tuple of `mindspore.Tensor` (one for the output of the embeddings + one for the output of each layer) of
            shape `(batch_size, sequence_length, hidden_size)`.

            Hidden-states of the model at the output of each layer plus the initial embedding outputs.
        entity_hidden_states (`tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`):
            Tuple of `mindspore.Tensor` (one for the output of the embeddings + one for the output of each layer) of
            shape `(batch_size, entity_length, hidden_size)`. Entity hidden-states of the model at the output of each
            layer plus the initial entity embedding outputs.
        attentions (`tuple(mindspore.Tensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`):
            Tuple of `mindspore.Tensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length,
            sequence_length)`.

            Attentions weights after the attention softmax, used to compute the weighted average in the self-attention
            heads.
    """

    loss: Optional[mindspore.Tensor] = None
    mlm_loss: Optional[mindspore.Tensor] = None
    mep_loss: Optional[mindspore.Tensor] = None
    logits: mindspore.Tensor = None
    entity_logits: mindspore.Tensor = None
    hidden_states: Optional[Tuple[mindspore.Tensor]] = None
    entity_hidden_states: Optional[Tuple[mindspore.Tensor, ...]] = None
    attentions: Optional[Tuple[mindspore.Tensor, ...]] = None

mindnlp.transformers.models.luke.modeling_luke.LukeModel

Bases: LukePreTrainedModel

Source code in mindnlp\transformers\models\luke\modeling_luke.py 
 907
 908
 909
 910
 911
 912
 913
 914
 915
 916
 917
 918
 919
 920
 921
 922
 923
 924
 925
 926
 927
 928
 929
 930
 931
 932
 933
 934
 935
 936
 937
 938
 939
 940
 941
 942
 943
 944
 945
 946
 947
 948
 949
 950
 951
 952
 953
 954
 955
 956
 957
 958
 959
 960
 961
 962
 963
 964
 965
 966
 967
 968
 969
 970
 971
 972
 973
 974
 975
 976
 977
 978
 979
 980
 981
 982
 983
 984
 985
 986
 987
 988
 989
 990
 991
 992
 993
 994
 995
 996
 997
 998
 999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
class LukeModel(LukePreTrainedModel):
    def __init__(self, config: LukeConfig, add_pooling_layer: bool = True):
        super().__init__(config)
        self.config = config

        self.embeddings = LukeEmbeddings(config)
        self.entity_embeddings = LukeEntityEmbeddings(config)
        self.encoder = LukeEncoder(config)

        self.pooler = LukePooler(config) if add_pooling_layer else None

        # Initialize weights and apply final processing
        self.post_init()

    def get_input_embeddings(self):
        return self.embeddings.word_embeddings

    def set_input_embeddings(self, value):
        self.embeddings.word_embeddings = value

    def get_entity_embeddings(self):
        return self.entity_embeddings.entity_embeddings

    def set_entity_embeddings(self, value):
        self.entity_embeddings.entity_embeddings = value

    def _prune_heads(self, heads_to_prune):
        raise NotImplementedError("LUKE does not support the pruning of attention heads")

    def forward(
        self,
        input_ids: Optional[mindspore.Tensor] = None,
        attention_mask: Optional[mindspore.Tensor] = None,
        token_type_ids: Optional[mindspore.Tensor] = None,
        position_ids: Optional[mindspore.Tensor] = None,
        entity_ids: Optional[mindspore.Tensor] = None,
        entity_attention_mask: Optional[mindspore.Tensor] = None,
        entity_token_type_ids: Optional[mindspore.Tensor] = None,
        entity_position_ids: Optional[mindspore.Tensor] = None,
        head_mask: Optional[mindspore.Tensor] = None,
        inputs_embeds: Optional[mindspore.Tensor] = None,
        output_attentions: Optional[bool] = None,
        output_hidden_states: Optional[bool] = None,
        return_dict: Optional[bool] = None,
    ) -> Union[Tuple, BaseLukeModelOutputWithPooling]:
        r"""

        Returns:

        Examples:

        ```python
        >>> from transformers import AutoTokenizer, LukeModel

        >>> tokenizer = AutoTokenizer.from_pretrained("studio-ousia/luke-base")
        >>> model = LukeModel.from_pretrained("studio-ousia/luke-base")
        # Compute the contextualized entity representation corresponding to the entity mention "Beyoncé"

        >>> text = "Beyoncé lives in Los Angeles."
        >>> entity_spans = [(0, 7)]  # character-based entity span corresponding to "Beyoncé"

        >>> encoding = tokenizer(text, entity_spans=entity_spans, add_prefix_space=True, return_tensors="ms")
        >>> outputs = model(**encoding)
        >>> word_last_hidden_state = outputs.last_hidden_state
        >>> entity_last_hidden_state = outputs.entity_last_hidden_state
        # Input Wikipedia entities to obtain enriched contextualized representations of word tokens

        >>> text = "Beyoncé lives in Los Angeles."
        >>> entities = [
        ...     "Beyoncé",
        ...     "Los Angeles",
        ... ]  # Wikipedia entity titles corresponding to the entity mentions "Beyoncé" and "Los Angeles"
        >>> entity_spans = [
        ...     (0, 7),
        ...     (17, 28),
        ... ]  # character-based entity spans corresponding to "Beyoncé" and "Los Angeles"

        >>> encoding = tokenizer(
        ...     text, entities=entities, entity_spans=entity_spans, add_prefix_space=True, return_tensors="ms"
        ... )
        >>> outputs = model(**encoding)
        >>> word_last_hidden_state = outputs.last_hidden_state
        >>> entity_last_hidden_state = outputs.entity_last_hidden_state
        ```"""
        output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions
        output_hidden_states = (
            output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states
        )
        return_dict = return_dict if return_dict is not None else self.config.use_return_dict

        if input_ids is not None and inputs_embeds is not None:
            raise ValueError("You cannot specify both input_ids and inputs_embeds at the same time")
        elif input_ids is not None:
            self.warn_if_padding_and_no_attention_mask(input_ids, attention_mask)
            input_shape = input_ids.shape
        elif inputs_embeds is not None:
            input_shape = inputs_embeds.shape[:-1]
        else:
            raise ValueError("You have to specify either input_ids or inputs_embeds")

        batch_size, seq_length = input_shape

        if attention_mask is None:
            attention_mask = ops.ones((batch_size, seq_length))
        if token_type_ids is None:
            token_type_ids = ops.zeros(input_shape, dtype=mindspore.int64)
        if entity_ids is not None:
            entity_seq_length = entity_ids.shape[1]
            if entity_attention_mask is None:
                entity_attention_mask = ops.ones((batch_size, entity_seq_length))
            if entity_token_type_ids is None:
                entity_token_type_ids = ops.zeros((batch_size, entity_seq_length), dtype=mindspore.int64)

        # Prepare head mask if needed
        # 1.0 in head_mask indicate we keep the head
        # attention_probs has shape bsz x n_heads x N x N
        # input head_mask has shape [num_heads] or [num_hidden_layers x num_heads]
        # and head_mask is converted to shape [num_hidden_layers x batch x num_heads x seq_length x seq_length]
        head_mask = self.get_head_mask(head_mask, self.config.num_hidden_layers)

        # First, compute word embeddings
        word_embedding_output = self.embeddings(
            input_ids=input_ids,
            position_ids=position_ids,
            token_type_ids=token_type_ids,
            inputs_embeds=inputs_embeds,
        )

        # Second, compute extended attention mask
        extended_attention_mask = self.get_extended_attention_mask(attention_mask, entity_attention_mask)

        # Third, compute entity embeddings and concatenate with word embeddings
        if entity_ids is None:
            entity_embedding_output = None
        else:
            entity_embedding_output = self.entity_embeddings(entity_ids, entity_position_ids, entity_token_type_ids)

        # Fourth, send embeddings through the model
        encoder_outputs = self.encoder(
            word_embedding_output,
            entity_embedding_output,
            attention_mask=extended_attention_mask,
            head_mask=head_mask,
            output_attentions=output_attentions,
            output_hidden_states=output_hidden_states,
            return_dict=return_dict,
        )

        # Fifth, get the output. LukeModel outputs the same as BertModel, namely sequence_output of shape (batch_size, seq_len, hidden_size)
        sequence_output = encoder_outputs[0]

        # Sixth, we compute the pooled_output, word_sequence_output and entity_sequence_output based on the sequence_output
        pooled_output = self.pooler(sequence_output) if self.pooler is not None else None

        if not return_dict:
            return (sequence_output, pooled_output) + encoder_outputs[1:]

        return BaseLukeModelOutputWithPooling(
            last_hidden_state=sequence_output,
            pooler_output=pooled_output,
            hidden_states=encoder_outputs.hidden_states,
            attentions=encoder_outputs.attentions,
            entity_last_hidden_state=encoder_outputs.entity_last_hidden_state,
            entity_hidden_states=encoder_outputs.entity_hidden_states,
        )

    def get_extended_attention_mask(
        self, word_attention_mask: mindspore.Tensor, entity_attention_mask: Optional[mindspore.Tensor]
    ):
        """
        Makes broadcastable attention and causal masks so that future and masked tokens are ignored.

        Arguments:
            word_attention_mask (`mindspore.Tensor`):
                Attention mask for word tokens with ones indicating tokens to attend to, zeros for tokens to ignore.
            entity_attention_mask (`mindspore.Tensor`, *optional*):
                Attention mask for entity tokens with ones indicating tokens to attend to, zeros for tokens to ignore.

        Returns:
            `mindspore.Tensor` The extended attention mask, with a the same dtype as `attention_mask.dtype`.
        """
        attention_mask = word_attention_mask
        if entity_attention_mask is not None:
            attention_mask = ops.cat([attention_mask, entity_attention_mask], dim=-1)

        if attention_mask.ndim == 3:
            extended_attention_mask = attention_mask[:, None, :, :]
        elif attention_mask.ndim == 2:
            extended_attention_mask = attention_mask[:, None, None, :]
        else:
            raise ValueError(f"Wrong shape for attention_mask (shape {attention_mask.shape})")

        extended_attention_mask = extended_attention_mask.to(dtype=self.dtype)  # fp16 compatibility
        extended_attention_mask = (1.0 - extended_attention_mask) * float(ops.finfo(self.dtype).min)
        return extended_attention_mask

mindnlp.transformers.models.luke.modeling_luke.LukeModel.forward(input_ids=None, attention_mask=None, token_type_ids=None, position_ids=None, entity_ids=None, entity_attention_mask=None, entity_token_type_ids=None, entity_position_ids=None, head_mask=None, inputs_embeds=None, output_attentions=None, output_hidden_states=None, return_dict=None)

Examples:

>>> from transformers import AutoTokenizer, LukeModel

>>> tokenizer = AutoTokenizer.from_pretrained("studio-ousia/luke-base")
>>> model = LukeModel.from_pretrained("studio-ousia/luke-base")
# Compute the contextualized entity representation corresponding to the entity mention "Beyoncé"

>>> text = "Beyoncé lives in Los Angeles."
>>> entity_spans = [(0, 7)]  # character-based entity span corresponding to "Beyoncé"

>>> encoding = tokenizer(text, entity_spans=entity_spans, add_prefix_space=True, return_tensors="ms")
>>> outputs = model(**encoding)
>>> word_last_hidden_state = outputs.last_hidden_state
>>> entity_last_hidden_state = outputs.entity_last_hidden_state
# Input Wikipedia entities to obtain enriched contextualized representations of word tokens

>>> text = "Beyoncé lives in Los Angeles."
>>> entities = [
...     "Beyoncé",
...     "Los Angeles",
... ]  # Wikipedia entity titles corresponding to the entity mentions "Beyoncé" and "Los Angeles"
>>> entity_spans = [
...     (0, 7),
...     (17, 28),
... ]  # character-based entity spans corresponding to "Beyoncé" and "Los Angeles"

>>> encoding = tokenizer(
...     text, entities=entities, entity_spans=entity_spans, add_prefix_space=True, return_tensors="ms"
... )
>>> outputs = model(**encoding)
>>> word_last_hidden_state = outputs.last_hidden_state
>>> entity_last_hidden_state = outputs.entity_last_hidden_state
Source code in mindnlp\transformers\models\luke\modeling_luke.py 
 936
 937
 938
 939
 940
 941
 942
 943
 944
 945
 946
 947
 948
 949
 950
 951
 952
 953
 954
 955
 956
 957
 958
 959
 960
 961
 962
 963
 964
 965
 966
 967
 968
 969
 970
 971
 972
 973
 974
 975
 976
 977
 978
 979
 980
 981
 982
 983
 984
 985
 986
 987
 988
 989
 990
 991
 992
 993
 994
 995
 996
 997
 998
 999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
def forward(
    self,
    input_ids: Optional[mindspore.Tensor] = None,
    attention_mask: Optional[mindspore.Tensor] = None,
    token_type_ids: Optional[mindspore.Tensor] = None,
    position_ids: Optional[mindspore.Tensor] = None,
    entity_ids: Optional[mindspore.Tensor] = None,
    entity_attention_mask: Optional[mindspore.Tensor] = None,
    entity_token_type_ids: Optional[mindspore.Tensor] = None,
    entity_position_ids: Optional[mindspore.Tensor] = None,
    head_mask: Optional[mindspore.Tensor] = None,
    inputs_embeds: Optional[mindspore.Tensor] = None,
    output_attentions: Optional[bool] = None,
    output_hidden_states: Optional[bool] = None,
    return_dict: Optional[bool] = None,
) -> Union[Tuple, BaseLukeModelOutputWithPooling]:
    r"""

    Returns:

    Examples:

    ```python
    >>> from transformers import AutoTokenizer, LukeModel

    >>> tokenizer = AutoTokenizer.from_pretrained("studio-ousia/luke-base")
    >>> model = LukeModel.from_pretrained("studio-ousia/luke-base")
    # Compute the contextualized entity representation corresponding to the entity mention "Beyoncé"

    >>> text = "Beyoncé lives in Los Angeles."
    >>> entity_spans = [(0, 7)]  # character-based entity span corresponding to "Beyoncé"

    >>> encoding = tokenizer(text, entity_spans=entity_spans, add_prefix_space=True, return_tensors="ms")
    >>> outputs = model(**encoding)
    >>> word_last_hidden_state = outputs.last_hidden_state
    >>> entity_last_hidden_state = outputs.entity_last_hidden_state
    # Input Wikipedia entities to obtain enriched contextualized representations of word tokens

    >>> text = "Beyoncé lives in Los Angeles."
    >>> entities = [
    ...     "Beyoncé",
    ...     "Los Angeles",
    ... ]  # Wikipedia entity titles corresponding to the entity mentions "Beyoncé" and "Los Angeles"
    >>> entity_spans = [
    ...     (0, 7),
    ...     (17, 28),
    ... ]  # character-based entity spans corresponding to "Beyoncé" and "Los Angeles"

    >>> encoding = tokenizer(
    ...     text, entities=entities, entity_spans=entity_spans, add_prefix_space=True, return_tensors="ms"
    ... )
    >>> outputs = model(**encoding)
    >>> word_last_hidden_state = outputs.last_hidden_state
    >>> entity_last_hidden_state = outputs.entity_last_hidden_state
    ```"""
    output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions
    output_hidden_states = (
        output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states
    )
    return_dict = return_dict if return_dict is not None else self.config.use_return_dict

    if input_ids is not None and inputs_embeds is not None:
        raise ValueError("You cannot specify both input_ids and inputs_embeds at the same time")
    elif input_ids is not None:
        self.warn_if_padding_and_no_attention_mask(input_ids, attention_mask)
        input_shape = input_ids.shape
    elif inputs_embeds is not None:
        input_shape = inputs_embeds.shape[:-1]
    else:
        raise ValueError("You have to specify either input_ids or inputs_embeds")

    batch_size, seq_length = input_shape

    if attention_mask is None:
        attention_mask = ops.ones((batch_size, seq_length))
    if token_type_ids is None:
        token_type_ids = ops.zeros(input_shape, dtype=mindspore.int64)
    if entity_ids is not None:
        entity_seq_length = entity_ids.shape[1]
        if entity_attention_mask is None:
            entity_attention_mask = ops.ones((batch_size, entity_seq_length))
        if entity_token_type_ids is None:
            entity_token_type_ids = ops.zeros((batch_size, entity_seq_length), dtype=mindspore.int64)

    # Prepare head mask if needed
    # 1.0 in head_mask indicate we keep the head
    # attention_probs has shape bsz x n_heads x N x N
    # input head_mask has shape [num_heads] or [num_hidden_layers x num_heads]
    # and head_mask is converted to shape [num_hidden_layers x batch x num_heads x seq_length x seq_length]
    head_mask = self.get_head_mask(head_mask, self.config.num_hidden_layers)

    # First, compute word embeddings
    word_embedding_output = self.embeddings(
        input_ids=input_ids,
        position_ids=position_ids,
        token_type_ids=token_type_ids,
        inputs_embeds=inputs_embeds,
    )

    # Second, compute extended attention mask
    extended_attention_mask = self.get_extended_attention_mask(attention_mask, entity_attention_mask)

    # Third, compute entity embeddings and concatenate with word embeddings
    if entity_ids is None:
        entity_embedding_output = None
    else:
        entity_embedding_output = self.entity_embeddings(entity_ids, entity_position_ids, entity_token_type_ids)

    # Fourth, send embeddings through the model
    encoder_outputs = self.encoder(
        word_embedding_output,
        entity_embedding_output,
        attention_mask=extended_attention_mask,
        head_mask=head_mask,
        output_attentions=output_attentions,
        output_hidden_states=output_hidden_states,
        return_dict=return_dict,
    )

    # Fifth, get the output. LukeModel outputs the same as BertModel, namely sequence_output of shape (batch_size, seq_len, hidden_size)
    sequence_output = encoder_outputs[0]

    # Sixth, we compute the pooled_output, word_sequence_output and entity_sequence_output based on the sequence_output
    pooled_output = self.pooler(sequence_output) if self.pooler is not None else None

    if not return_dict:
        return (sequence_output, pooled_output) + encoder_outputs[1:]

    return BaseLukeModelOutputWithPooling(
        last_hidden_state=sequence_output,
        pooler_output=pooled_output,
        hidden_states=encoder_outputs.hidden_states,
        attentions=encoder_outputs.attentions,
        entity_last_hidden_state=encoder_outputs.entity_last_hidden_state,
        entity_hidden_states=encoder_outputs.entity_hidden_states,
    )

mindnlp.transformers.models.luke.modeling_luke.LukeModel.get_extended_attention_mask(word_attention_mask, entity_attention_mask)

Makes broadcastable attention and causal masks so that future and masked tokens are ignored.

PARAMETER DESCRIPTION
word_attention_mask

Attention mask for word tokens with ones indicating tokens to attend to, zeros for tokens to ignore.

TYPE: `mindspore.Tensor`

entity_attention_mask

Attention mask for entity tokens with ones indicating tokens to attend to, zeros for tokens to ignore.

TYPE: `mindspore.Tensor`, *optional*

RETURNS DESCRIPTION

mindspore.Tensor The extended attention mask, with a the same dtype as attention_mask.dtype.
Source code in mindnlp\transformers\models\luke\modeling_luke.py 
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
def get_extended_attention_mask(
    self, word_attention_mask: mindspore.Tensor, entity_attention_mask: Optional[mindspore.Tensor]
):
    """
    Makes broadcastable attention and causal masks so that future and masked tokens are ignored.

    Arguments:
        word_attention_mask (`mindspore.Tensor`):
            Attention mask for word tokens with ones indicating tokens to attend to, zeros for tokens to ignore.
        entity_attention_mask (`mindspore.Tensor`, *optional*):
            Attention mask for entity tokens with ones indicating tokens to attend to, zeros for tokens to ignore.

    Returns:
        `mindspore.Tensor` The extended attention mask, with a the same dtype as `attention_mask.dtype`.
    """
    attention_mask = word_attention_mask
    if entity_attention_mask is not None:
        attention_mask = ops.cat([attention_mask, entity_attention_mask], dim=-1)

    if attention_mask.ndim == 3:
        extended_attention_mask = attention_mask[:, None, :, :]
    elif attention_mask.ndim == 2:
        extended_attention_mask = attention_mask[:, None, None, :]
    else:
        raise ValueError(f"Wrong shape for attention_mask (shape {attention_mask.shape})")

    extended_attention_mask = extended_attention_mask.to(dtype=self.dtype)  # fp16 compatibility
    extended_attention_mask = (1.0 - extended_attention_mask) * float(ops.finfo(self.dtype).min)
    return extended_attention_mask

mindnlp.transformers.models.luke.modeling_luke.LukeMultipleChoiceModelOutput dataclass

Bases: ModelOutput

Outputs of multiple choice models.

PARAMETER DESCRIPTION
loss

Classification loss.

TYPE: `mindspore.Tensor` of shape *(1,)*, *optional*, returned when `labels` is provided DEFAULT: None

logits

num_choices is the second dimension of the input tensors. (see input_ids above).

Classification scores (before SoftMax).

TYPE: `mindspore.Tensor` of shape `(batch_size, num_choices)` DEFAULT: None

hidden_states

Tuple of mindspore.Tensor (one for the output of the embeddings, if the model has an embedding layer, + one for the output of each layer) of shape (batch_size, sequence_length, hidden_size).

Hidden-states of the model at the output of each layer plus the optional initial embedding outputs.

TYPE: `tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True` DEFAULT: None

entity_hidden_states

Tuple of mindspore.Tensor (one for the output of the embeddings + one for the output of each layer) of shape (batch_size, entity_length, hidden_size). Entity hidden-states of the model at the output of each layer plus the initial entity embedding outputs.

TYPE: `tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True` DEFAULT: None

attentions

Tuple of mindspore.Tensor (one for each layer) of shape (batch_size, num_heads, sequence_length, sequence_length).

Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads.

TYPE: `tuple(mindspore.Tensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True` DEFAULT: None

Source code in mindnlp\transformers\models\luke\modeling_luke.py 
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
@dataclass
class LukeMultipleChoiceModelOutput(ModelOutput):
    """
    Outputs of multiple choice models.

    Args:
        loss (`mindspore.Tensor` of shape *(1,)*, *optional*, returned when `labels` is provided):
            Classification loss.
        logits (`mindspore.Tensor` of shape `(batch_size, num_choices)`):
            *num_choices* is the second dimension of the input tensors. (see *input_ids* above).

            Classification scores (before SoftMax).
        hidden_states (`tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`):
            Tuple of `mindspore.Tensor` (one for the output of the embeddings, if the model has an embedding layer, +
            one for the output of each layer) of shape `(batch_size, sequence_length, hidden_size)`.

            Hidden-states of the model at the output of each layer plus the optional initial embedding outputs.
        entity_hidden_states (`tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`):
            Tuple of `mindspore.Tensor` (one for the output of the embeddings + one for the output of each layer) of
            shape `(batch_size, entity_length, hidden_size)`. Entity hidden-states of the model at the output of each
            layer plus the initial entity embedding outputs.
        attentions (`tuple(mindspore.Tensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`):
            Tuple of `mindspore.Tensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length,
            sequence_length)`.

            Attentions weights after the attention softmax, used to compute the weighted average in the self-attention
            heads.
    """

    loss: Optional[mindspore.Tensor] = None
    logits: mindspore.Tensor = None
    hidden_states: Optional[Tuple[mindspore.Tensor, ...]] = None
    entity_hidden_states: Optional[Tuple[mindspore.Tensor, ...]] = None
    attentions: Optional[Tuple[mindspore.Tensor, ...]] = None

mindnlp.transformers.models.luke.modeling_luke.LukePreTrainedModel

Bases: PreTrainedModel

An abstract class to handle weights initialization and a simple interface for downloading and loading pretrained models.

Source code in mindnlp\transformers\models\luke\modeling_luke.py 
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
class LukePreTrainedModel(PreTrainedModel):
    """
    An abstract class to handle weights initialization and a simple interface for downloading and loading pretrained
    models.
    """

    config_class = LukeConfig
    base_model_prefix = "luke"
    supports_gradient_checkpointing = True
    _no_split_modules = ["LukeAttention", "LukeEntityEmbeddings"]

    def _init_weights(self, module: nn.Module):
        """Initialize the weights"""
        if isinstance(module, nn.Linear):
            nn.init.normal_(module.weight, mean=0.0, std=self.config.initializer_range)
            if module.bias is not None:
                nn.init.zeros_(module.bias)
        elif isinstance(module, nn.Embedding):
            if module.embedding_dim == 1:  # embedding for bias parameters
                nn.init.zeros_(module.weight)
            else:
                nn.init.normal_(module.weight, mean=0.0, std=self.config.initializer_range)
            if module.padding_idx is not None:
                module.weight.data[module.padding_idx] = 0
        elif isinstance(module, nn.LayerNorm):
            nn.init.zeros_(module.bias)
            nn.init.ones_(module.weight)

mindnlp.transformers.models.luke.modeling_luke.LukeQuestionAnsweringModelOutput dataclass

Bases: ModelOutput

Outputs of question answering models.

PARAMETER DESCRIPTION
loss

Total span extraction loss is the sum of a Cross-Entropy for the start and end positions.

TYPE: `mindspore.Tensor` of shape `(1,)`, *optional*, returned when `labels` is provided DEFAULT: None

start_logits

Span-start scores (before SoftMax).

TYPE: `mindspore.Tensor` of shape `(batch_size, sequence_length)` DEFAULT: None

end_logits

Span-end scores (before SoftMax).

TYPE: `mindspore.Tensor` of shape `(batch_size, sequence_length)` DEFAULT: None

hidden_states

Tuple of mindspore.Tensor (one for the output of the embeddings, if the model has an embedding layer, + one for the output of each layer) of shape (batch_size, sequence_length, hidden_size).

Hidden-states of the model at the output of each layer plus the optional initial embedding outputs.

TYPE: `tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True` DEFAULT: None

entity_hidden_states

Tuple of mindspore.Tensor (one for the output of the embeddings + one for the output of each layer) of shape (batch_size, entity_length, hidden_size). Entity hidden-states of the model at the output of each layer plus the initial entity embedding outputs.

TYPE: `tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True` DEFAULT: None

attentions

Tuple of mindspore.Tensor (one for each layer) of shape (batch_size, num_heads, sequence_length, sequence_length).

Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads.

TYPE: `tuple(mindspore.Tensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True` DEFAULT: None

Source code in mindnlp\transformers\models\luke\modeling_luke.py 
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
@dataclass
class LukeQuestionAnsweringModelOutput(ModelOutput):
    """
    Outputs of question answering models.

    Args:
        loss (`mindspore.Tensor` of shape `(1,)`, *optional*, returned when `labels` is provided):
            Total span extraction loss is the sum of a Cross-Entropy for the start and end positions.
        start_logits (`mindspore.Tensor` of shape `(batch_size, sequence_length)`):
            Span-start scores (before SoftMax).
        end_logits (`mindspore.Tensor` of shape `(batch_size, sequence_length)`):
            Span-end scores (before SoftMax).
        hidden_states (`tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`):
            Tuple of `mindspore.Tensor` (one for the output of the embeddings, if the model has an embedding layer, +
            one for the output of each layer) of shape `(batch_size, sequence_length, hidden_size)`.

            Hidden-states of the model at the output of each layer plus the optional initial embedding outputs.
        entity_hidden_states (`tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`):
            Tuple of `mindspore.Tensor` (one for the output of the embeddings + one for the output of each layer) of
            shape `(batch_size, entity_length, hidden_size)`. Entity hidden-states of the model at the output of each
            layer plus the initial entity embedding outputs.
        attentions (`tuple(mindspore.Tensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`):
            Tuple of `mindspore.Tensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length,
            sequence_length)`.

            Attentions weights after the attention softmax, used to compute the weighted average in the self-attention
            heads.
    """

    loss: Optional[mindspore.Tensor] = None
    start_logits: mindspore.Tensor = None
    end_logits: mindspore.Tensor = None
    hidden_states: Optional[Tuple[mindspore.Tensor, ...]] = None
    entity_hidden_states: Optional[Tuple[mindspore.Tensor, ...]] = None
    attentions: Optional[Tuple[mindspore.Tensor, ...]] = None

mindnlp.transformers.models.luke.modeling_luke.LukeSequenceClassifierOutput dataclass

Bases: ModelOutput

Outputs of sentence classification models.

PARAMETER DESCRIPTION
loss

Classification (or regression if config.num_labels==1) loss.

TYPE: `mindspore.Tensor` of shape `(1,)`, *optional*, returned when `labels` is provided DEFAULT: None

logits

Classification (or regression if config.num_labels==1) scores (before SoftMax).

TYPE: `mindspore.Tensor` of shape `(batch_size, config.num_labels)` DEFAULT: None

hidden_states

Tuple of mindspore.Tensor (one for the output of the embeddings, if the model has an embedding layer, + one for the output of each layer) of shape (batch_size, sequence_length, hidden_size).

Hidden-states of the model at the output of each layer plus the optional initial embedding outputs.

TYPE: `tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True` DEFAULT: None

entity_hidden_states

Tuple of mindspore.Tensor (one for the output of the embeddings + one for the output of each layer) of shape (batch_size, entity_length, hidden_size). Entity hidden-states of the model at the output of each layer plus the initial entity embedding outputs.

TYPE: `tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True` DEFAULT: None

attentions

Tuple of mindspore.Tensor (one for each layer) of shape (batch_size, num_heads, sequence_length, sequence_length).

Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads.

TYPE: `tuple(mindspore.Tensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True` DEFAULT: None

Source code in mindnlp\transformers\models\luke\modeling_luke.py 
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
@dataclass
class LukeSequenceClassifierOutput(ModelOutput):
    """
    Outputs of sentence classification models.

    Args:
        loss (`mindspore.Tensor` of shape `(1,)`, *optional*, returned when `labels` is provided):
            Classification (or regression if config.num_labels==1) loss.
        logits (`mindspore.Tensor` of shape `(batch_size, config.num_labels)`):
            Classification (or regression if config.num_labels==1) scores (before SoftMax).
        hidden_states (`tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`):
            Tuple of `mindspore.Tensor` (one for the output of the embeddings, if the model has an embedding layer, +
            one for the output of each layer) of shape `(batch_size, sequence_length, hidden_size)`.

            Hidden-states of the model at the output of each layer plus the optional initial embedding outputs.
        entity_hidden_states (`tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`):
            Tuple of `mindspore.Tensor` (one for the output of the embeddings + one for the output of each layer) of
            shape `(batch_size, entity_length, hidden_size)`. Entity hidden-states of the model at the output of each
            layer plus the initial entity embedding outputs.
        attentions (`tuple(mindspore.Tensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`):
            Tuple of `mindspore.Tensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length,
            sequence_length)`.

            Attentions weights after the attention softmax, used to compute the weighted average in the self-attention
            heads.
    """

    loss: Optional[mindspore.Tensor] = None
    logits: mindspore.Tensor = None
    hidden_states: Optional[Tuple[mindspore.Tensor, ...]] = None
    entity_hidden_states: Optional[Tuple[mindspore.Tensor, ...]] = None
    attentions: Optional[Tuple[mindspore.Tensor, ...]] = None

mindnlp.transformers.models.luke.modeling_luke.LukeTokenClassifierOutput dataclass

Bases: ModelOutput

Base class for outputs of token classification models.

PARAMETER DESCRIPTION
loss

Classification loss.

TYPE: `mindspore.Tensor` of shape `(1,)`, *optional*, returned when `labels` is provided) DEFAULT: None

logits

Classification scores (before SoftMax).

TYPE: `mindspore.Tensor` of shape `(batch_size, sequence_length, config.num_labels)` DEFAULT: None

hidden_states

Tuple of mindspore.Tensor (one for the output of the embeddings, if the model has an embedding layer, + one for the output of each layer) of shape (batch_size, sequence_length, hidden_size).

Hidden-states of the model at the output of each layer plus the optional initial embedding outputs.

TYPE: `tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True` DEFAULT: None

entity_hidden_states

Tuple of mindspore.Tensor (one for the output of the embeddings + one for the output of each layer) of shape (batch_size, entity_length, hidden_size). Entity hidden-states of the model at the output of each layer plus the initial entity embedding outputs.

TYPE: `tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True` DEFAULT: None

attentions

Tuple of mindspore.Tensor (one for each layer) of shape (batch_size, num_heads, sequence_length, sequence_length).

Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads.

TYPE: `tuple(mindspore.Tensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True` DEFAULT: None

Source code in mindnlp\transformers\models\luke\modeling_luke.py 
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
@dataclass
class LukeTokenClassifierOutput(ModelOutput):
    """
    Base class for outputs of token classification models.

    Args:
        loss (`mindspore.Tensor` of shape `(1,)`, *optional*, returned when `labels` is provided) :
            Classification loss.
        logits (`mindspore.Tensor` of shape `(batch_size, sequence_length, config.num_labels)`):
            Classification scores (before SoftMax).
        hidden_states (`tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`):
            Tuple of `mindspore.Tensor` (one for the output of the embeddings, if the model has an embedding layer, +
            one for the output of each layer) of shape `(batch_size, sequence_length, hidden_size)`.

            Hidden-states of the model at the output of each layer plus the optional initial embedding outputs.
        entity_hidden_states (`tuple(mindspore.Tensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`):
            Tuple of `mindspore.Tensor` (one for the output of the embeddings + one for the output of each layer) of
            shape `(batch_size, entity_length, hidden_size)`. Entity hidden-states of the model at the output of each
            layer plus the initial entity embedding outputs.
        attentions (`tuple(mindspore.Tensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`):
            Tuple of `mindspore.Tensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length,
            sequence_length)`.

            Attentions weights after the attention softmax, used to compute the weighted average in the self-attention
            heads.
    """

    loss: Optional[mindspore.Tensor] = None
    logits: mindspore.Tensor = None
    hidden_states: Optional[Tuple[mindspore.Tensor, ...]] = None
    entity_hidden_states: Optional[Tuple[mindspore.Tensor, ...]] = None
    attentions: Optional[Tuple[mindspore.Tensor, ...]] = None

mindnlp.transformers.models.luke.modeling_luke.create_position_ids_from_input_ids(input_ids, padding_idx)

Replace non-padding symbols with their position numbers. Position numbers begin at padding_idx+1. Padding symbols are ignored. This is modified from fairseq's utils.make_positions.

PARAMETER DESCRIPTION
x

mindspore.Tensor x:
Source code in mindnlp\transformers\models\luke\modeling_luke.py 
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
def create_position_ids_from_input_ids(input_ids, padding_idx):
    """
    Replace non-padding symbols with their position numbers. Position numbers begin at padding_idx+1. Padding symbols
    are ignored. This is modified from fairseq's `utils.make_positions`.

    Args:
        x: mindspore.Tensor x:

    Returns: mindspore.Tensor
    """
    # The series of casts and type-conversions here are carefully balanced to both work with ONNX export and XLA.
    mask = input_ids.ne(padding_idx).int()
    incremental_indices = (ops.cumsum(mask, dim=1).type_as(mask)) * mask
    return incremental_indices.long() + padding_idx

mindnlp.transformers.models.luke.configuration_luke

LUKE configuration

mindnlp.transformers.models.luke.configuration_luke.LukeConfig

Bases: PretrainedConfig

This is the configuration class to store the configuration of a [LukeModel]. It is used to instantiate a LUKE model according to the specified arguments, defining the model architecture. Instantiating a configuration with the defaults will yield a similar configuration to that of the LUKE studio-ousia/luke-base architecture.

Configuration objects inherit from [PretrainedConfig] and can be used to control the model outputs. Read the documentation from [PretrainedConfig] for more information.

PARAMETER DESCRIPTION
vocab_size

Vocabulary size of the LUKE model. Defines the number of different tokens that can be represented by the inputs_ids passed when calling [LukeModel].

TYPE: `int`, *optional*, defaults to 50267 DEFAULT: 50267

entity_vocab_size

Entity vocabulary size of the LUKE model. Defines the number of different entities that can be represented by the entity_ids passed when calling [LukeModel].

TYPE: `int`, *optional*, defaults to 500000 DEFAULT: 500000

hidden_size

Dimensionality of the encoder layers and the pooler layer.

TYPE: `int`, *optional*, defaults to 768 DEFAULT: 768

entity_emb_size

The number of dimensions of the entity embedding.

TYPE: `int`, *optional*, defaults to 256 DEFAULT: 256

num_hidden_layers

Number of hidden layers in the Transformer encoder.

TYPE: `int`, *optional*, defaults to 12 DEFAULT: 12

num_attention_heads

Number of attention heads for each attention layer in the Transformer encoder.

TYPE: `int`, *optional*, defaults to 12 DEFAULT: 12

intermediate_size

Dimensionality of the "intermediate" (often named feed-forward) layer in the Transformer encoder.

TYPE: `int`, *optional*, defaults to 3072 DEFAULT: 3072

hidden_act

The non-linear activation function (function or string) in the encoder and pooler. If string, "gelu", "relu", "silu" and "gelu_new" are supported.

TYPE: `str` or `Callable`, *optional*, defaults to `"gelu"` DEFAULT: 'gelu'

hidden_dropout_prob

The dropout probability for all fully connected layers in the embeddings, encoder, and pooler.

TYPE: `float`, *optional*, defaults to 0.1 DEFAULT: 0.1

attention_probs_dropout_prob

The dropout ratio for the attention probabilities.

TYPE: `float`, *optional*, defaults to 0.1 DEFAULT: 0.1

max_position_embeddings

The maximum sequence length that this model might ever be used with. Typically set this to something large just in case (e.g., 512 or 1024 or 2048).

TYPE: `int`, *optional*, defaults to 512 DEFAULT: 512

type_vocab_size

The vocabulary size of the token_type_ids passed when calling [LukeModel].

TYPE: `int`, *optional*, defaults to 2 DEFAULT: 2

initializer_range

The standard deviation of the truncated_normal_initializer for initializing all weight matrices.

TYPE: `float`, *optional*, defaults to 0.02 DEFAULT: 0.02

layer_norm_eps

The epsilon used by the layer normalization layers.

TYPE: `float`, *optional*, defaults to 1e-12 DEFAULT: 1e-12

use_entity_aware_attention

Whether or not the model should use the entity-aware self-attention mechanism proposed in LUKE: Deep Contextualized Entity Representations with Entity-aware Self-attention (Yamada et al.).

TYPE: `bool`, *optional*, defaults to `True` DEFAULT: True

classifier_dropout

The dropout ratio for the classification head.

TYPE: `float`, *optional* DEFAULT: None

pad_token_id

Padding token id.

TYPE: `int`, *optional*, defaults to 1 DEFAULT: 1

bos_token_id

Beginning of stream token id.

TYPE: `int`, *optional*, defaults to 0 DEFAULT: 0

eos_token_id

End of stream token id.

TYPE: `int`, *optional*, defaults to 2 DEFAULT: 2 

>>> from transformers import LukeConfig, LukeModel

>>> # Initializing a LUKE configuration
>>> configuration = LukeConfig()

>>> # Initializing a model from the configuration
>>> model = LukeModel(configuration)

>>> # Accessing the model configuration
>>> configuration = model.config
Source code in mindnlp\transformers\models\luke\configuration_luke.py 
 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
class LukeConfig(PretrainedConfig):
    r"""
    This is the configuration class to store the configuration of a [`LukeModel`]. It is used to instantiate a LUKE
    model according to the specified arguments, defining the model architecture. Instantiating a configuration with the
    defaults will yield a similar configuration to that of the LUKE
    [studio-ousia/luke-base](https://huggingface.co/studio-ousia/luke-base) architecture.

    Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the
    documentation from [`PretrainedConfig`] for more information.


    Args:
        vocab_size (`int`, *optional*, defaults to 50267):
            Vocabulary size of the LUKE model. Defines the number of different tokens that can be represented by the
            `inputs_ids` passed when calling [`LukeModel`].
        entity_vocab_size (`int`, *optional*, defaults to 500000):
            Entity vocabulary size of the LUKE model. Defines the number of different entities that can be represented
            by the `entity_ids` passed when calling [`LukeModel`].
        hidden_size (`int`, *optional*, defaults to 768):
            Dimensionality of the encoder layers and the pooler layer.
        entity_emb_size (`int`, *optional*, defaults to 256):
            The number of dimensions of the entity embedding.
        num_hidden_layers (`int`, *optional*, defaults to 12):
            Number of hidden layers in the Transformer encoder.
        num_attention_heads (`int`, *optional*, defaults to 12):
            Number of attention heads for each attention layer in the Transformer encoder.
        intermediate_size (`int`, *optional*, defaults to 3072):
            Dimensionality of the "intermediate" (often named feed-forward) layer in the Transformer encoder.
        hidden_act (`str` or `Callable`, *optional*, defaults to `"gelu"`):
            The non-linear activation function (function or string) in the encoder and pooler. If string, `"gelu"`,
            `"relu"`, `"silu"` and `"gelu_new"` are supported.
        hidden_dropout_prob (`float`, *optional*, defaults to 0.1):
            The dropout probability for all fully connected layers in the embeddings, encoder, and pooler.
        attention_probs_dropout_prob (`float`, *optional*, defaults to 0.1):
            The dropout ratio for the attention probabilities.
        max_position_embeddings (`int`, *optional*, defaults to 512):
            The maximum sequence length that this model might ever be used with. Typically set this to something large
            just in case (e.g., 512 or 1024 or 2048).
        type_vocab_size (`int`, *optional*, defaults to 2):
            The vocabulary size of the `token_type_ids` passed when calling [`LukeModel`].
        initializer_range (`float`, *optional*, defaults to 0.02):
            The standard deviation of the truncated_normal_initializer for initializing all weight matrices.
        layer_norm_eps (`float`, *optional*, defaults to 1e-12):
            The epsilon used by the layer normalization layers.
        use_entity_aware_attention (`bool`, *optional*, defaults to `True`):
            Whether or not the model should use the entity-aware self-attention mechanism proposed in [LUKE: Deep
            Contextualized Entity Representations with Entity-aware Self-attention (Yamada et
            al.)](https://arxiv.org/abs/2010.01057).
        classifier_dropout (`float`, *optional*):
            The dropout ratio for the classification head.
        pad_token_id (`int`, *optional*, defaults to 1):
            Padding token id.
        bos_token_id (`int`, *optional*, defaults to 0):
            Beginning of stream token id.
        eos_token_id (`int`, *optional*, defaults to 2):
            End of stream token id.

    Examples:

    ```python
    >>> from transformers import LukeConfig, LukeModel

    >>> # Initializing a LUKE configuration
    >>> configuration = LukeConfig()

    >>> # Initializing a model from the configuration
    >>> model = LukeModel(configuration)

    >>> # Accessing the model configuration
    >>> configuration = model.config
    ```"""

    model_type = "luke"

    def __init__(
        self,
        vocab_size=50267,
        entity_vocab_size=500000,
        hidden_size=768,
        entity_emb_size=256,
        num_hidden_layers=12,
        num_attention_heads=12,
        intermediate_size=3072,
        hidden_act="gelu",
        hidden_dropout_prob=0.1,
        attention_probs_dropout_prob=0.1,
        max_position_embeddings=512,
        type_vocab_size=2,
        initializer_range=0.02,
        layer_norm_eps=1e-12,
        use_entity_aware_attention=True,
        classifier_dropout=None,
        pad_token_id=1,
        bos_token_id=0,
        eos_token_id=2,
        **kwargs,
    ):
        """Constructs LukeConfig."""
        super().__init__(pad_token_id=pad_token_id, bos_token_id=bos_token_id, eos_token_id=eos_token_id, **kwargs)

        self.vocab_size = vocab_size
        self.entity_vocab_size = entity_vocab_size
        self.hidden_size = hidden_size
        self.entity_emb_size = entity_emb_size
        self.num_hidden_layers = num_hidden_layers
        self.num_attention_heads = num_attention_heads
        self.hidden_act = hidden_act
        self.intermediate_size = intermediate_size
        self.hidden_dropout_prob = hidden_dropout_prob
        self.attention_probs_dropout_prob = attention_probs_dropout_prob
        self.max_position_embeddings = max_position_embeddings
        self.type_vocab_size = type_vocab_size
        self.initializer_range = initializer_range
        self.layer_norm_eps = layer_norm_eps
        self.use_entity_aware_attention = use_entity_aware_attention
        self.classifier_dropout = classifier_dropout

mindnlp.transformers.models.luke.configuration_luke.LukeConfig.__init__(vocab_size=50267, entity_vocab_size=500000, hidden_size=768, entity_emb_size=256, num_hidden_layers=12, num_attention_heads=12, intermediate_size=3072, hidden_act='gelu', hidden_dropout_prob=0.1, attention_probs_dropout_prob=0.1, max_position_embeddings=512, type_vocab_size=2, initializer_range=0.02, layer_norm_eps=1e-12, use_entity_aware_attention=True, classifier_dropout=None, pad_token_id=1, bos_token_id=0, eos_token_id=2, **kwargs)

Constructs LukeConfig.

Source code in mindnlp\transformers\models\luke\configuration_luke.py 
 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
def __init__(
    self,
    vocab_size=50267,
    entity_vocab_size=500000,
    hidden_size=768,
    entity_emb_size=256,
    num_hidden_layers=12,
    num_attention_heads=12,
    intermediate_size=3072,
    hidden_act="gelu",
    hidden_dropout_prob=0.1,
    attention_probs_dropout_prob=0.1,
    max_position_embeddings=512,
    type_vocab_size=2,
    initializer_range=0.02,
    layer_norm_eps=1e-12,
    use_entity_aware_attention=True,
    classifier_dropout=None,
    pad_token_id=1,
    bos_token_id=0,
    eos_token_id=2,
    **kwargs,
):
    """Constructs LukeConfig."""
    super().__init__(pad_token_id=pad_token_id, bos_token_id=bos_token_id, eos_token_id=eos_token_id, **kwargs)

    self.vocab_size = vocab_size
    self.entity_vocab_size = entity_vocab_size
    self.hidden_size = hidden_size
    self.entity_emb_size = entity_emb_size
    self.num_hidden_layers = num_hidden_layers
    self.num_attention_heads = num_attention_heads
    self.hidden_act = hidden_act
    self.intermediate_size = intermediate_size
    self.hidden_dropout_prob = hidden_dropout_prob
    self.attention_probs_dropout_prob = attention_probs_dropout_prob
    self.max_position_embeddings = max_position_embeddings
    self.type_vocab_size = type_vocab_size
    self.initializer_range = initializer_range
    self.layer_norm_eps = layer_norm_eps
    self.use_entity_aware_attention = use_entity_aware_attention
    self.classifier_dropout = classifier_dropout

mindnlp.transformers.models.luke.tokenization_luke

Tokenization classes for LUKE.

mindnlp.transformers.models.luke.tokenization_luke.LukeTokenizer

Bases: PreTrainedTokenizer

Constructs a LUKE tokenizer, derived from the GPT-2 tokenizer, using byte-level Byte-Pair-Encoding.

This tokenizer has been trained to treat spaces like parts of the tokens (a bit like sentencepiece) so a word will be encoded differently whether it is at the beginning of the sentence (without space) or not:

Example 
>>> from transformers import LukeTokenizer
...
>>> tokenizer = LukeTokenizer.from_pretrained("studio-ousia/luke-base")
>>> tokenizer("Hello world")["input_ids"]
[0, 31414, 232, 2]
>>> tokenizer(" Hello world")["input_ids"]
[0, 20920, 232, 2]

You can get around that behavior by passing add_prefix_space=True when instantiating this tokenizer or when you call it on some text, but since the model was not pretrained this way, it might yield a decrease in performance.

When used with is_split_into_words=True, this tokenizer will add a space before each word (even the first one).

This tokenizer inherits from [PreTrainedTokenizer] which contains most of the main methods. Users should refer to this superclass for more information regarding those methods. It also creates entity sequences, namely entity_ids, entity_attention_mask, entity_token_type_ids, and entity_position_ids to be used by the LUKE model.

PARAMETER DESCRIPTION
vocab_file

Path to the vocabulary file.

TYPE: `str`

merges_file

Path to the merges file.

TYPE: `str`

entity_vocab_file

Path to the entity vocabulary file.

TYPE: `str`

task

Task for which you want to prepare sequences. One of "entity_classification", "entity_pair_classification", or "entity_span_classification". If you specify this argument, the entity sequence is automatically created based on the given entity span(s).

TYPE: `str`, *optional* DEFAULT: None

max_entity_length

The maximum length of entity_ids.

TYPE: `int`, *optional*, defaults to 32 DEFAULT: 32

max_mention_length

The maximum number of tokens inside an entity span.

TYPE: `int`, *optional*, defaults to 30 DEFAULT: 30

entity_token_1

The special token used to represent an entity span in a word token sequence. This token is only used when task is set to "entity_classification" or "entity_pair_classification".

TYPE: `str`, *optional*, defaults to `<ent>` DEFAULT: '<ent>'

entity_token_2

The special token used to represent an entity span in a word token sequence. This token is only used when task is set to "entity_pair_classification".

TYPE: `str`, *optional*, defaults to `<ent2>` DEFAULT: '<ent2>'

errors

Paradigm to follow when decoding bytes to UTF-8. See bytes.decode for more information.

TYPE: `str`, *optional*, defaults to `"replace"` DEFAULT: 'replace'

bos_token

The beginning of sequence token that was used during pretraining. Can be used a sequence classifier token.

When building a sequence using special tokens, this is not the token that is used for the beginning of sequence. The token used is the cls_token.

TYPE: `str`, *optional*, defaults to `"<s>"` DEFAULT: '<s>'

eos_token

The end of sequence token.

When building a sequence using special tokens, this is not the token that is used for the end of sequence. The token used is the sep_token.

TYPE: `str`, *optional*, defaults to `"</s>"` DEFAULT: '</s>'

sep_token

The separator token, which is used when building a sequence from multiple sequences, e.g. two sequences for sequence classification or for a text and a question for question answering. It is also used as the last token of a sequence built with special tokens.

TYPE: `str`, *optional*, defaults to `"</s>"` DEFAULT: '</s>'

cls_token

The classifier token which is used when doing sequence classification (classification of the whole sequence instead of per-token classification). It is the first token of the sequence when built with special tokens.

TYPE: `str`, *optional*, defaults to `"<s>"` DEFAULT: '<s>'

unk_token

The unknown token. A token that is not in the vocabulary cannot be converted to an ID and is set to be this token instead.

TYPE: `str`, *optional*, defaults to `"<unk>"` DEFAULT: '<unk>'

pad_token

The token used for padding, for example when batching sequences of different lengths.

TYPE: `str`, *optional*, defaults to `"<pad>"` DEFAULT: '<pad>'

mask_token

The token used for masking values. This is the token used when training this model with masked language modeling. This is the token which the model will try to predict.

TYPE: `str`, *optional*, defaults to `"<mask>"` DEFAULT: '<mask>'

add_prefix_space

Whether or not to add an initial space to the input. This allows to treat the leading word just as any other word. (LUKE tokenizer detect beginning of words by the preceding space).

TYPE: `bool`, *optional*, defaults to `False` DEFAULT: False

Source code in mindnlp\transformers\models\luke\tokenization_luke.py 
 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
 673
 674
 675
 676
 677
 678
 679
 680
 681
 682
 683
 684
 685
 686
 687
 688
 689
 690
 691
 692
 693
 694
 695
 696
 697
 698
 699
 700
 701
 702
 703
 704
 705
 706
 707
 708
 709
 710
 711
 712
 713
 714
 715
 716
 717
 718
 719
 720
 721
 722
 723
 724
 725
 726
 727
 728
 729
 730
 731
 732
 733
 734
 735
 736
 737
 738
 739
 740
 741
 742
 743
 744
 745
 746
 747
 748
 749
 750
 751
 752
 753
 754
 755
 756
 757
 758
 759
 760
 761
 762
 763
 764
 765
 766
 767
 768
 769
 770
 771
 772
 773
 774
 775
 776
 777
 778
 779
 780
 781
 782
 783
 784
 785
 786
 787
 788
 789
 790
 791
 792
 793
 794
 795
 796
 797
 798
 799
 800
 801
 802
 803
 804
 805
 806
 807
 808
 809
 810
 811
 812
 813
 814
 815
 816
 817
 818
 819
 820
 821
 822
 823
 824
 825
 826
 827
 828
 829
 830
 831
 832
 833
 834
 835
 836
 837
 838
 839
 840
 841
 842
 843
 844
 845
 846
 847
 848
 849
 850
 851
 852
 853
 854
 855
 856
 857
 858
 859
 860
 861
 862
 863
 864
 865
 866
 867
 868
 869
 870
 871
 872
 873
 874
 875
 876
 877
 878
 879
 880
 881
 882
 883
 884
 885
 886
 887
 888
 889
 890
 891
 892
 893
 894
 895
 896
 897
 898
 899
 900
 901
 902
 903
 904
 905
 906
 907
 908
 909
 910
 911
 912
 913
 914
 915
 916
 917
 918
 919
 920
 921
 922
 923
 924
 925
 926
 927
 928
 929
 930
 931
 932
 933
 934
 935
 936
 937
 938
 939
 940
 941
 942
 943
 944
 945
 946
 947
 948
 949
 950
 951
 952
 953
 954
 955
 956
 957
 958
 959
 960
 961
 962
 963
 964
 965
 966
 967
 968
 969
 970
 971
 972
 973
 974
 975
 976
 977
 978
 979
 980
 981
 982
 983
 984
 985
 986
 987
 988
 989
 990
 991
 992
 993
 994
 995
 996
 997
 998
 999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
1845
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
1882
1883
1884
1885
1886
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
class LukeTokenizer(PreTrainedTokenizer):
    """
    Constructs a LUKE tokenizer, derived from the GPT-2 tokenizer, using byte-level Byte-Pair-Encoding.

    This tokenizer has been trained to treat spaces like parts of the tokens (a bit like sentencepiece) so a word will
    be encoded differently whether it is at the beginning of the sentence (without space) or not:

    Example:
        ```python
        >>> from transformers import LukeTokenizer
        ...
        >>> tokenizer = LukeTokenizer.from_pretrained("studio-ousia/luke-base")
        >>> tokenizer("Hello world")["input_ids"]
        [0, 31414, 232, 2]
        >>> tokenizer(" Hello world")["input_ids"]
        [0, 20920, 232, 2]
        ```

    You can get around that behavior by passing `add_prefix_space=True` when instantiating this tokenizer or when you
    call it on some text, but since the model was not pretrained this way, it might yield a decrease in performance.

    <Tip>

    When used with `is_split_into_words=True`, this tokenizer will add a space before each word (even the first one).

    </Tip>

    This tokenizer inherits from [`PreTrainedTokenizer`] which contains most of the main methods. Users should refer to
    this superclass for more information regarding those methods. It also creates entity sequences, namely
    `entity_ids`, `entity_attention_mask`, `entity_token_type_ids`, and `entity_position_ids` to be used by the LUKE
    model.

    Args:
        vocab_file (`str`):
            Path to the vocabulary file.
        merges_file (`str`):
            Path to the merges file.
        entity_vocab_file (`str`):
            Path to the entity vocabulary file.
        task (`str`, *optional*):
            Task for which you want to prepare sequences. One of `"entity_classification"`,
            `"entity_pair_classification"`, or `"entity_span_classification"`. If you specify this argument, the entity
            sequence is automatically created based on the given entity span(s).
        max_entity_length (`int`, *optional*, defaults to 32):
            The maximum length of `entity_ids`.
        max_mention_length (`int`, *optional*, defaults to 30):
            The maximum number of tokens inside an entity span.
        entity_token_1 (`str`, *optional*, defaults to `<ent>`):
            The special token used to represent an entity span in a word token sequence. This token is only used when
            `task` is set to `"entity_classification"` or `"entity_pair_classification"`.
        entity_token_2 (`str`, *optional*, defaults to `<ent2>`):
            The special token used to represent an entity span in a word token sequence. This token is only used when
            `task` is set to `"entity_pair_classification"`.
        errors (`str`, *optional*, defaults to `"replace"`):
            Paradigm to follow when decoding bytes to UTF-8. See
            [bytes.decode](https://docs.python.org/3/library/stdtypes.html#bytes.decode) for more information.
        bos_token (`str`, *optional*, defaults to `"<s>"`):
            The beginning of sequence token that was used during pretraining. Can be used a sequence classifier token.

            <Tip>

            When building a sequence using special tokens, this is not the token that is used for the beginning of
            sequence. The token used is the `cls_token`.

            </Tip>

        eos_token (`str`, *optional*, defaults to `"</s>"`):
            The end of sequence token.

            <Tip>

            When building a sequence using special tokens, this is not the token that is used for the end of sequence.
            The token used is the `sep_token`.

            </Tip>

        sep_token (`str`, *optional*, defaults to `"</s>"`):
            The separator token, which is used when building a sequence from multiple sequences, e.g. two sequences for
            sequence classification or for a text and a question for question answering. It is also used as the last
            token of a sequence built with special tokens.
        cls_token (`str`, *optional*, defaults to `"<s>"`):
            The classifier token which is used when doing sequence classification (classification of the whole sequence
            instead of per-token classification). It is the first token of the sequence when built with special tokens.
        unk_token (`str`, *optional*, defaults to `"<unk>"`):
            The unknown token. A token that is not in the vocabulary cannot be converted to an ID and is set to be this
            token instead.
        pad_token (`str`, *optional*, defaults to `"<pad>"`):
            The token used for padding, for example when batching sequences of different lengths.
        mask_token (`str`, *optional*, defaults to `"<mask>"`):
            The token used for masking values. This is the token used when training this model with masked language
            modeling. This is the token which the model will try to predict.
        add_prefix_space (`bool`, *optional*, defaults to `False`):
            Whether or not to add an initial space to the input. This allows to treat the leading word just as any
            other word. (LUKE tokenizer detect beginning of words by the preceding space).
    """
    vocab_files_names = VOCAB_FILES_NAMES
    pretrained_vocab_files_map = PRETRAINED_VOCAB_FILES_MAP
    max_model_input_sizes = PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES
    model_input_names = ["input_ids", "attention_mask"]

    def __init__(
        self,
        vocab_file,
        merges_file,
        entity_vocab_file,
        task=None,
        max_entity_length=32,
        max_mention_length=30,
        entity_token_1="<ent>",
        entity_token_2="<ent2>",
        entity_unk_token="[UNK]",
        entity_pad_token="[PAD]",
        entity_mask_token="[MASK]",
        entity_mask2_token="[MASK2]",
        errors="replace",
        bos_token="<s>",
        eos_token="</s>",
        sep_token="</s>",
        cls_token="<s>",
        unk_token="<unk>",
        pad_token="<pad>",
        mask_token="<mask>",
        add_prefix_space=False,
        **kwargs,
    ):
        """Initialize the LukeTokenizer class.

        This method initializes an instance of the LukeTokenizer class. It takes the following parameters:

        Args:
            self: The instance of the class.
            vocab_file (str): The path to the vocabulary file.
            merges_file (str): The path to the merges file.
            entity_vocab_file (str): The path to the entity vocabulary file.
            task (str, optional): The task for which the tokenizer is used. Defaults to None.
            max_entity_length (int, optional): The maximum length of the entity. Defaults to 32.
            max_mention_length (int, optional): The maximum length of the mention. Defaults to 30.
            entity_token_1 (str, optional): The first entity token. Defaults to '<ent>'.
            entity_token_2 (str, optional): The second entity token. Defaults to '<ent2>'.
            entity_unk_token (str, optional): The unknown entity token. Defaults to '[UNK]'.
            entity_pad_token (str, optional): The padding entity token. Defaults to '[PAD]'.
            entity_mask_token (str, optional): The masked entity token. Defaults to '[MASK]'.
            entity_mask2_token (str, optional): The second masked entity token. Defaults to '[MASK2]'.
            errors (str, optional): The error handling strategy. Defaults to 'replace'.
            bos_token (str, optional): The beginning of sentence token. Defaults to '<s>'.
            eos_token (str, optional): The end of sentence token. Defaults to '</s>'.
            sep_token (str, optional): The separator token. Defaults to '</s>'.
            cls_token (str, optional): The classification token. Defaults to '<s>'.
            unk_token (str, optional): The unknown token. Defaults to '<unk>'.
            pad_token (str, optional): The padding token. Defaults to '<pad>'.
            mask_token (str, optional): The masked token. Defaults to '<mask>'.
            add_prefix_space (bool, optional): Whether to add space before the token. Defaults to False.

        Returns:
            None

        Raises:
            ValueError: If the specified entity special token is not found in the entity vocabulary file.
            ValueError: If the task is not supported. Select task from ['entity_classification',
                'entity_pair_classification', 'entity_span_classification'] only.
        """
        bos_token = AddedToken(bos_token, lstrip=False, rstrip=False) if isinstance(bos_token, str) else bos_token
        eos_token = AddedToken(eos_token, lstrip=False, rstrip=False) if isinstance(eos_token, str) else eos_token
        sep_token = AddedToken(sep_token, lstrip=False, rstrip=False) if isinstance(sep_token, str) else sep_token
        cls_token = AddedToken(cls_token, lstrip=False, rstrip=False) if isinstance(cls_token, str) else cls_token
        unk_token = AddedToken(unk_token, lstrip=False, rstrip=False) if isinstance(unk_token, str) else unk_token
        pad_token = AddedToken(pad_token, lstrip=False, rstrip=False) if isinstance(pad_token, str) else pad_token

        # Mask token behave like a normal word, i.e. include the space before it
        mask_token = AddedToken(mask_token, lstrip=True, rstrip=False) if isinstance(mask_token, str) else mask_token

        with open(vocab_file, encoding="utf-8") as vocab_handle:
            self.encoder = json.load(vocab_handle)
        self.decoder = {v: k for k, v in self.encoder.items()}
        self.errors = errors  # how to handle errors in decoding
        self.byte_encoder = bytes_to_unicode()
        self.byte_decoder = {v: k for k, v in self.byte_encoder.items()}
        with open(merges_file, encoding="utf-8") as merges_handle:
            bpe_merges = merges_handle.read().split("\n")[1:-1]
        bpe_merges = [tuple(merge.split()) for merge in bpe_merges]
        self.bpe_ranks = dict(zip(bpe_merges, range(len(bpe_merges))))
        self.cache = {}
        self.add_prefix_space = add_prefix_space

        # Should have added re.IGNORECASE so BPE merges can happen for capitalized versions of contractions
        self.pat = re.compile(r"""'s|'t|'re|'ve|'m|'ll|'d| ?\p{L}+| ?\p{N}+| ?[^\s\p{L}\p{N}]+|\s+(?!\S)|\s+""")

        # we add 2 special tokens for downstream tasks
        # for more information about lstrip and rstrip, see https://github.com/huggingface/transformers/pull/2778
        entity_token_1 = (
            AddedToken(entity_token_1, lstrip=False, rstrip=False)
            if isinstance(entity_token_1, str)
            else entity_token_1
        )
        entity_token_2 = (
            AddedToken(entity_token_2, lstrip=False, rstrip=False)
            if isinstance(entity_token_2, str)
            else entity_token_2
        )
        kwargs["additional_special_tokens"] = kwargs.get("additional_special_tokens", [])
        kwargs["additional_special_tokens"] += [entity_token_1, entity_token_2]

        with open(entity_vocab_file, encoding="utf-8") as entity_vocab_handle:
            self.entity_vocab = json.load(entity_vocab_handle)
        for entity_special_token in [entity_unk_token, entity_pad_token, entity_mask_token, entity_mask2_token]:
            if entity_special_token not in self.entity_vocab:
                raise ValueError(
                    f"Specified entity special token ``{entity_special_token}`` is not found in entity_vocab. "
                    f"Probably an incorrect entity vocab file is loaded: {entity_vocab_file}."
                )
        self.entity_unk_token_id = self.entity_vocab[entity_unk_token]
        self.entity_pad_token_id = self.entity_vocab[entity_pad_token]
        self.entity_mask_token_id = self.entity_vocab[entity_mask_token]
        self.entity_mask2_token_id = self.entity_vocab[entity_mask2_token]

        self.task = task
        if task is None or task == "entity_span_classification":
            self.max_entity_length = max_entity_length
        elif task == "entity_classification":
            self.max_entity_length = 1
        elif task == "entity_pair_classification":
            self.max_entity_length = 2
        else:
            raise ValueError(
                f"Task {task} not supported. Select task from ['entity_classification', 'entity_pair_classification',"
                " 'entity_span_classification'] only."
            )

        self.max_mention_length = max_mention_length

        super().__init__(
            errors=errors,
            bos_token=bos_token,
            eos_token=eos_token,
            unk_token=unk_token,
            sep_token=sep_token,
            cls_token=cls_token,
            pad_token=pad_token,
            mask_token=mask_token,
            add_prefix_space=add_prefix_space,
            task=task,
            max_entity_length=32,
            max_mention_length=30,
            entity_token_1="<ent>",
            entity_token_2="<ent2>",
            entity_unk_token=entity_unk_token,
            entity_pad_token=entity_pad_token,
            entity_mask_token=entity_mask_token,
            entity_mask2_token=entity_mask2_token,
            **kwargs,
        )

    @property
    # Copied from transformers.models.roberta.tokenization_roberta.RobertaTokenizer.vocab_size with Roberta->Luke, RoBERTa->LUKE
    def vocab_size(self):
        """
        Returns the size of the vocabulary.

        Args:
            self (LukeTokenizer): The instance of the LukeTokenizer class.

        Returns:
            int: The number of items in the encoder, representing the size of the vocabulary.

        Raises:
            None.
        """
        return len(self.encoder)

    # Copied from transformers.models.roberta.tokenization_roberta.RobertaTokenizer.get_vocab with Roberta->Luke, RoBERTa->LUKE
    def get_vocab(self):
        """
        Retrieves the vocabulary dictionary for the 'LukeTokenizer' class.

        Args:
            self: An instance of the 'LukeTokenizer' class.

        Returns:
            dict: A dictionary containing the vocabulary of the tokenizer. The keys are the tokens
                and the values are their corresponding IDs.

        Raises:
            None.
        """
        vocab = dict(self.encoder).copy()
        vocab.update(self.added_tokens_encoder)
        return vocab

    # Copied from transformers.models.roberta.tokenization_roberta.RobertaTokenizer.bpe with Roberta->Luke, RoBERTa->LUKE
    def bpe(self, token):
        """
        This method 'bpe' in the class 'LukeTokenizer' performs Byte Pair Encoding (BPE) on the input token.

        Args:
            self (LukeTokenizer): The instance of the LukeTokenizer class.
            token (str): The input token to be processed using Byte Pair Encoding.

        Returns:
            str: The processed token after applying Byte Pair Encoding.

        Raises:
            ValueError: If the input token is empty.
            TypeError: If the input token is not a string.
        """
        if token in self.cache:
            return self.cache[token]
        word = tuple(token)
        pairs = get_pairs(word)

        if not pairs:
            return token

        while True:
            bigram = min(pairs, key=lambda pair: self.bpe_ranks.get(pair, float("inf")))
            if bigram not in self.bpe_ranks:
                break
            first, second = bigram
            new_word = []
            i = 0
            while i < len(word):
                try:
                    j = word.index(first, i)
                except ValueError:
                    new_word.extend(word[i:])
                    break
                else:
                    new_word.extend(word[i:j])
                    i = j

                if word[i] == first and i < len(word) - 1 and word[i + 1] == second:
                    new_word.append(first + second)
                    i += 2
                else:
                    new_word.append(word[i])
                    i += 1
            new_word = tuple(new_word)
            word = new_word
            if len(word) == 1:
                break
            pairs = get_pairs(word)
        word = " ".join(word)
        self.cache[token] = word
        return word

    # Copied from transformers.models.roberta.tokenization_roberta.RobertaTokenizer._tokenize with Roberta->Luke, RoBERTa->LUKE
    def _tokenize(self, text):
        """Tokenize a string."""
        bpe_tokens = []
        for token in re.findall(self.pat, text):
            token = "".join(
                self.byte_encoder[b] for b in token.encode("utf-8")
            )  # Maps all our bytes to unicode strings, avoiding control tokens of the BPE (spaces in our case)
            bpe_tokens.extend(bpe_token for bpe_token in self.bpe(token).split(" "))
        return bpe_tokens

    # Copied from transformers.models.roberta.tokenization_roberta.RobertaTokenizer._convert_token_to_id with Roberta->Luke, RoBERTa->LUKE
    def _convert_token_to_id(self, token):
        """Converts a token (str) in an id using the vocab."""
        return self.encoder.get(token, self.encoder.get(self.unk_token))

    # Copied from transformers.models.roberta.tokenization_roberta.RobertaTokenizer._convert_id_to_token with Roberta->Luke, RoBERTa->LUKE
    def _convert_id_to_token(self, index):
        """Converts an index (integer) in a token (str) using the vocab."""
        return self.decoder.get(index)

    # Copied from transformers.models.roberta.tokenization_roberta.RobertaTokenizer.convert_tokens_to_string with Roberta->Luke, RoBERTa->LUKE
    def convert_tokens_to_string(self, tokens):
        """Converts a sequence of tokens (string) in a single string."""
        text = "".join(tokens)
        text = bytearray([self.byte_decoder[c] for c in text]).decode("utf-8", errors=self.errors)
        return text

    # Copied from transformers.models.roberta.tokenization_roberta.RobertaTokenizer.build_inputs_with_special_tokens with Roberta->Luke, RoBERTa->LUKE
    def build_inputs_with_special_tokens(
        self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None
    ) -> List[int]:
        """
        Build model inputs from a sequence or a pair of sequence for sequence classification tasks by concatenating and
        adding special tokens. A LUKE sequence has the following format:

        - single sequence: `<s> X </s>`
        - pair of sequences: `<s> A </s></s> B </s>`

        Args:
            token_ids_0 (`List[int]`):
                List of IDs to which the special tokens will be added.
            token_ids_1 (`List[int]`, *optional*):
                Optional second list of IDs for sequence pairs.

        Returns:
            `List[int]`: List of [input IDs](../glossary#input-ids) with the appropriate special tokens.
        """
        if token_ids_1 is None:
            return [self.cls_token_id] + token_ids_0 + [self.sep_token_id]
        cls = [self.cls_token_id]
        sep = [self.sep_token_id]
        return cls + token_ids_0 + sep + sep + token_ids_1 + sep

    # Copied from transformers.models.roberta.tokenization_roberta.RobertaTokenizer.get_special_tokens_mask with Roberta->Luke, RoBERTa->LUKE
    def get_special_tokens_mask(
        self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None, already_has_special_tokens: bool = False
    ) -> List[int]:
        """
        Retrieve sequence ids from a token list that has no special tokens added. This method is called when adding
        special tokens using the tokenizer `prepare_for_model` method.

        Args:
            token_ids_0 (`List[int]`):
                List of IDs.
            token_ids_1 (`List[int]`, *optional*):
                Optional second list of IDs for sequence pairs.
            already_has_special_tokens (`bool`, *optional*, defaults to `False`):
                Whether or not the token list is already formatted with special tokens for the model.

        Returns:
            `List[int]`: A list of integers in the range [0, 1]: 1 for a special token, 0 for a sequence token.
        """
        if already_has_special_tokens:
            return super().get_special_tokens_mask(
                token_ids_0=token_ids_0, token_ids_1=token_ids_1, already_has_special_tokens=True
            )

        if token_ids_1 is None:
            return [1] + ([0] * len(token_ids_0)) + [1]
        return [1] + ([0] * len(token_ids_0)) + [1, 1] + ([0] * len(token_ids_1)) + [1]

    # Copied from transformers.models.roberta.tokenization_roberta.RobertaTokenizer.create_token_type_ids_from_sequences with Roberta->Luke, RoBERTa->LUKE
    def create_token_type_ids_from_sequences(
        self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None
    ) -> List[int]:
        """
        Create a mask from the two sequences passed to be used in a sequence-pair classification task. LUKE does not
        make use of token type ids, therefore a list of zeros is returned.

        Args:
            token_ids_0 (`List[int]`):
                List of IDs.
            token_ids_1 (`List[int]`, *optional*):
                Optional second list of IDs for sequence pairs.

        Returns:
            `List[int]`: List of zeros.
        """
        sep = [self.sep_token_id]
        cls = [self.cls_token_id]

        if token_ids_1 is None:
            return len(cls + token_ids_0 + sep) * [0]
        return len(cls + token_ids_0 + sep + sep + token_ids_1 + sep) * [0]

    # Copied from transformers.models.roberta.tokenization_roberta.RobertaTokenizer.prepare_for_tokenization with Roberta->Luke, RoBERTa->LUKE
    def prepare_for_tokenization(self, text, is_split_into_words=False, **kwargs):
        """
        Prepares the input text for tokenization by adding a prefix space if necessary.

        Args:
            self (LukeTokenizer): An instance of the LukeTokenizer class.
            text (str): The input text to be tokenized.
            is_split_into_words (bool): A flag indicating if the input text is already split into words.
                Defaults to False.

        Returns:
            None: The method modifies the input text in-place.

        Raises:
            None.
        """
        add_prefix_space = kwargs.pop("add_prefix_space", self.add_prefix_space)
        if (is_split_into_words or add_prefix_space) and (len(text) > 0 and not text[0].isspace()):
            text = " " + text
        return (text, kwargs)

    def __call__(
        self,
        text: Union[TextInput, List[TextInput]],
        text_pair: Optional[Union[TextInput, List[TextInput]]] = None,
        entity_spans: Optional[Union[EntitySpanInput, List[EntitySpanInput]]] = None,
        entity_spans_pair: Optional[Union[EntitySpanInput, List[EntitySpanInput]]] = None,
        entities: Optional[Union[EntityInput, List[EntityInput]]] = None,
        entities_pair: Optional[Union[EntityInput, List[EntityInput]]] = None,
        add_special_tokens: bool = True,
        padding: Union[bool, str, PaddingStrategy] = False,
        truncation: Union[bool, str, TruncationStrategy] = None,
        max_length: Optional[int] = None,
        max_entity_length: Optional[int] = None,
        stride: int = 0,
        is_split_into_words: Optional[bool] = False,
        pad_to_multiple_of: Optional[int] = None,
        return_tensors: Optional[Union[str, TensorType]] = None,
        return_token_type_ids: Optional[bool] = None,
        return_attention_mask: Optional[bool] = None,
        return_overflowing_tokens: bool = False,
        return_special_tokens_mask: bool = False,
        return_offsets_mapping: bool = False,
        return_length: bool = False,
        verbose: bool = True,
        **kwargs,
    ) -> BatchEncoding:
        """
        Main method to tokenize and prepare for the model one or several sequence(s) or one or several pair(s) of
        sequences, depending on the task you want to prepare them for.

        Args:
            text (`str`, `List[str]`, `List[List[str]]`):
                The sequence or batch of sequences to be encoded. Each sequence must be a string. Note that this
                tokenizer does not support tokenization based on pretokenized strings.
            text_pair (`str`, `List[str]`, `List[List[str]]`):
                The sequence or batch of sequences to be encoded. Each sequence must be a string. Note that this
                tokenizer does not support tokenization based on pretokenized strings.
            entity_spans (`List[Tuple[int, int]]`, `List[List[Tuple[int, int]]]`, *optional*):
                The sequence or batch of sequences of entity spans to be encoded. Each sequence consists of tuples each
                with two integers denoting character-based start and end positions of entities. If you specify
                `"entity_classification"` or `"entity_pair_classification"` as the `task` argument in the forwardor,
                the length of each sequence must be 1 or 2, respectively. If you specify `entities`, the length of each
                sequence must be equal to the length of each sequence of `entities`.
            entity_spans_pair (`List[Tuple[int, int]]`, `List[List[Tuple[int, int]]]`, *optional*):
                The sequence or batch of sequences of entity spans to be encoded. Each sequence consists of tuples each
                with two integers denoting character-based start and end positions of entities. If you specify the
                `task` argument in the forwardor, this argument is ignored. If you specify `entities_pair`, the
                length of each sequence must be equal to the length of each sequence of `entities_pair`.
            entities (`List[str]`, `List[List[str]]`, *optional*):
                The sequence or batch of sequences of entities to be encoded. Each sequence consists of strings
                representing entities, i.e., special entities (e.g., [MASK]) or entity titles of Wikipedia (e.g., Los
                Angeles). This argument is ignored if you specify the `task` argument in the forwardor. The length of
                each sequence must be equal to the length of each sequence of `entity_spans`. If you specify
                `entity_spans` without specifying this argument, the entity sequence or the batch of entity sequences
                is automatically forwarded by filling it with the [MASK] entity.
            entities_pair (`List[str]`, `List[List[str]]`, *optional*):
                The sequence or batch of sequences of entities to be encoded. Each sequence consists of strings
                representing entities, i.e., special entities (e.g., [MASK]) or entity titles of Wikipedia (e.g., Los
                Angeles). This argument is ignored if you specify the `task` argument in the forwardor. The length of
                each sequence must be equal to the length of each sequence of `entity_spans_pair`. If you specify
                `entity_spans_pair` without specifying this argument, the entity sequence or the batch of entity
                sequences is automatically forwarded by filling it with the [MASK] entity.
            max_entity_length (`int`, *optional*):
                The maximum length of `entity_ids`.
        """
        # Input type checking for clearer error
        is_valid_single_text = isinstance(text, str)
        is_valid_batch_text = isinstance(text, (list, tuple)) and (len(text) == 0 or (isinstance(text[0], str)))
        if not (is_valid_single_text or is_valid_batch_text):
            raise ValueError("text input must be of type `str` (single example) or `List[str]` (batch).")

        is_valid_single_text_pair = isinstance(text_pair, str)
        is_valid_batch_text_pair = isinstance(text_pair, (list, tuple)) and (
            len(text_pair) == 0 or isinstance(text_pair[0], str)
        )
        if not (text_pair is None or is_valid_single_text_pair or is_valid_batch_text_pair):
            raise ValueError("text_pair input must be of type `str` (single example) or `List[str]` (batch).")

        is_batched = bool(isinstance(text, (list, tuple)))

        if is_batched:
            batch_text_or_text_pairs = list(zip(text, text_pair)) if text_pair is not None else text
            if entities is None:
                batch_entities_or_entities_pairs = None
            else:
                batch_entities_or_entities_pairs = (
                    list(zip(entities, entities_pair)) if entities_pair is not None else entities
                )

            if entity_spans is None:
                batch_entity_spans_or_entity_spans_pairs = None
            else:
                batch_entity_spans_or_entity_spans_pairs = (
                    list(zip(entity_spans, entity_spans_pair)) if entity_spans_pair is not None else entity_spans
                )

            return self.batch_encode_plus(
                batch_text_or_text_pairs=batch_text_or_text_pairs,
                batch_entity_spans_or_entity_spans_pairs=batch_entity_spans_or_entity_spans_pairs,
                batch_entities_or_entities_pairs=batch_entities_or_entities_pairs,
                add_special_tokens=add_special_tokens,
                padding=padding,
                truncation=truncation,
                max_length=max_length,
                max_entity_length=max_entity_length,
                stride=stride,
                is_split_into_words=is_split_into_words,
                pad_to_multiple_of=pad_to_multiple_of,
                return_tensors=return_tensors,
                return_token_type_ids=return_token_type_ids,
                return_attention_mask=return_attention_mask,
                return_overflowing_tokens=return_overflowing_tokens,
                return_special_tokens_mask=return_special_tokens_mask,
                return_offsets_mapping=return_offsets_mapping,
                return_length=return_length,
                verbose=verbose,
                **kwargs,
            )
        return self.encode_plus(
            text=text,
            text_pair=text_pair,
            entity_spans=entity_spans,
            entity_spans_pair=entity_spans_pair,
            entities=entities,
            entities_pair=entities_pair,
            add_special_tokens=add_special_tokens,
            padding=padding,
            truncation=truncation,
            max_length=max_length,
            max_entity_length=max_entity_length,
            stride=stride,
            is_split_into_words=is_split_into_words,
            pad_to_multiple_of=pad_to_multiple_of,
            return_tensors=return_tensors,
            return_token_type_ids=return_token_type_ids,
            return_attention_mask=return_attention_mask,
            return_overflowing_tokens=return_overflowing_tokens,
            return_special_tokens_mask=return_special_tokens_mask,
            return_offsets_mapping=return_offsets_mapping,
            return_length=return_length,
            verbose=verbose,
            **kwargs,
        )

    def _encode_plus(
        self,
        text: TextInput,
        text_pair: Optional[TextInput] = None,
        entity_spans: Optional[EntitySpanInput] = None,
        entity_spans_pair: Optional[EntitySpanInput] = None,
        entities: Optional[EntityInput] = None,
        entities_pair: Optional[EntityInput] = None,
        add_special_tokens: bool = True,
        padding_strategy: PaddingStrategy = PaddingStrategy.DO_NOT_PAD,
        truncation_strategy: TruncationStrategy = TruncationStrategy.DO_NOT_TRUNCATE,
        max_length: Optional[int] = None,
        max_entity_length: Optional[int] = None,
        stride: int = 0,
        is_split_into_words: Optional[bool] = False,
        pad_to_multiple_of: Optional[int] = None,
        return_tensors: Optional[Union[str, TensorType]] = None,
        return_token_type_ids: Optional[bool] = None,
        return_attention_mask: Optional[bool] = None,
        return_overflowing_tokens: bool = False,
        return_special_tokens_mask: bool = False,
        return_offsets_mapping: bool = False,
        return_length: bool = False,
        verbose: bool = True,
        **kwargs,
    ) -> BatchEncoding:
        """Encodes the inputs for the LukeTokenizer.

        Args:
            self (LukeTokenizer): The instance of the LukeTokenizer class.
            text (TextInput): The input text to be encoded. It can be a single sentence or a sequence of sentences.
            text_pair (Optional[TextInput], optional): The second input text to be encoded.
                It can be a single sentence or a sequence of sentences. Defaults to None.
            entity_spans (Optional[EntitySpanInput], optional): The input entity spans to be encoded. Defaults to None.
            entity_spans_pair (Optional[EntitySpanInput], optional): The second input entity spans to be encoded.
                Defaults to None.
            entities (Optional[EntityInput], optional): The input entities to be encoded. Defaults to None.
            entities_pair (Optional[EntityInput], optional): The second input entities to be encoded. Defaults to None.
            add_special_tokens (bool, optional): Whether to add special tokens to the encoded inputs. Defaults to True.
            padding_strategy (PaddingStrategy, optional): The strategy to use for padding.
                Defaults to PaddingStrategy.DO_NOT_PAD.
            truncation_strategy (TruncationStrategy, optional): The strategy to use for truncation.
                Defaults to TruncationStrategy.DO_NOT_TRUNCATE.
            max_length (Optional[int], optional): The maximum sequence length after encoding. Defaults to None.
            max_entity_length (Optional[int], optional): The maximum entity span length after encoding. Defaults to None.
            stride (int, optional): The stride to use for overflowing tokens. Defaults to 0.
            is_split_into_words (Optional[bool], optional): Whether the input text is already split into words.
                Defaults to False.
            pad_to_multiple_of (Optional[int], optional): The padding length will be a multiple of this value.
                Defaults to None.
            return_tensors (Optional[Union[str, TensorType]], optional): The type of tensors to return. Defaults to None.
            return_token_type_ids (Optional[bool], optional): Whether to return token type IDs. Defaults to None.
            return_attention_mask (Optional[bool], optional): Whether to return attention masks. Defaults to None.
            return_overflowing_tokens (bool, optional): Whether to return the overflowing tokens. Defaults to False.
            return_special_tokens_mask (bool, optional): Whether to return the special tokens mask. Defaults to False.
            return_offsets_mapping (bool, optional): Whether to return the offsets mapping. Defaults to False.
            return_length (bool, optional): Whether to return the length of the encoded inputs. Defaults to False.
            verbose (bool, optional): Whether to print verbose logs. Defaults to True.
            **kwargs: Additional keyword arguments.

        Returns:
            BatchEncoding: The encoded inputs as a BatchEncoding object.

        Raises:
            NotImplementedError: If return_offsets_mapping is requested.
            NotImplementedError: If is_split_into_words is True and not supported by the tokenizer.
        """
        if return_offsets_mapping:
            raise NotImplementedError(
                "return_offset_mapping is not available when using Python tokenizers. "
                "To use this feature, change your tokenizer to one deriving from "
                "transformers.PreTrainedTokenizerFast. "
                "More information on available tokenizers at "
                "https://github.com/huggingface/transformers/pull/2674"
            )

        if is_split_into_words:
            raise NotImplementedError("is_split_into_words is not supported in this tokenizer.")

        (
            first_ids,
            second_ids,
            first_entity_ids,
            second_entity_ids,
            first_entity_token_spans,
            second_entity_token_spans,
        ) = self._create_input_sequence(
            text=text,
            text_pair=text_pair,
            entities=entities,
            entities_pair=entities_pair,
            entity_spans=entity_spans,
            entity_spans_pair=entity_spans_pair,
            **kwargs,
        )

        # prepare_for_model will create the attention_mask and token_type_ids
        return self.prepare_for_model(
            first_ids,
            pair_ids=second_ids,
            entity_ids=first_entity_ids,
            pair_entity_ids=second_entity_ids,
            entity_token_spans=first_entity_token_spans,
            pair_entity_token_spans=second_entity_token_spans,
            add_special_tokens=add_special_tokens,
            padding=padding_strategy.value,
            truncation=truncation_strategy.value,
            max_length=max_length,
            max_entity_length=max_entity_length,
            stride=stride,
            pad_to_multiple_of=pad_to_multiple_of,
            return_tensors=return_tensors,
            prepend_batch_axis=True,
            return_attention_mask=return_attention_mask,
            return_token_type_ids=return_token_type_ids,
            return_overflowing_tokens=return_overflowing_tokens,
            return_special_tokens_mask=return_special_tokens_mask,
            return_length=return_length,
            verbose=verbose,
        )

    def _batch_encode_plus(
        self,
        batch_text_or_text_pairs: Union[List[TextInput], List[TextInputPair]],
        batch_entity_spans_or_entity_spans_pairs: Optional[
            Union[List[EntitySpanInput], List[Tuple[EntitySpanInput, EntitySpanInput]]]
        ] = None,
        batch_entities_or_entities_pairs: Optional[
            Union[List[EntityInput], List[Tuple[EntityInput, EntityInput]]]
        ] = None,
        add_special_tokens: bool = True,
        padding_strategy: PaddingStrategy = PaddingStrategy.DO_NOT_PAD,
        truncation_strategy: TruncationStrategy = TruncationStrategy.DO_NOT_TRUNCATE,
        max_length: Optional[int] = None,
        max_entity_length: Optional[int] = None,
        stride: int = 0,
        is_split_into_words: Optional[bool] = False,
        pad_to_multiple_of: Optional[int] = None,
        return_tensors: Optional[Union[str, TensorType]] = None,
        return_token_type_ids: Optional[bool] = None,
        return_attention_mask: Optional[bool] = None,
        return_overflowing_tokens: bool = False,
        return_special_tokens_mask: bool = False,
        return_offsets_mapping: bool = False,
        return_length: bool = False,
        verbose: bool = True,
        **kwargs,
    ) -> BatchEncoding:
        """
        Performs batch encoding of text and entity inputs for the LukeTokenizer class.

        Args:
            self (LukeTokenizer): The LukeTokenizer instance.
            batch_text_or_text_pairs (Union[List[TextInput], List[TextInputPair]]):
                A list of text inputs or text pairs to be encoded.
            batch_entity_spans_or_entity_spans_pairs (Optional[Union[List[EntitySpanInput], List[Tuple[EntitySpanInput, EntitySpanInput]]]]):
                A list of entity span inputs or entity span input pairs to be encoded. Defaults to None.
            batch_entities_or_entities_pairs (Optional[Union[List[EntityInput], List[Tuple[EntityInput, EntityInput]]]]):
                A list of entity inputs or entity input pairs to be encoded. Defaults to None.
            add_special_tokens (bool): Whether to add special tokens to the encoded inputs. Defaults to True.
            padding_strategy (PaddingStrategy): The strategy to use for padding. Defaults to PaddingStrategy.DO_NOT_PAD.
            truncation_strategy (TruncationStrategy): The strategy to use for truncation.
                Defaults to TruncationStrategy.DO_NOT_TRUNCATE.
            max_length (Optional[int]): The maximum length of the encoded inputs. Defaults to None.
            max_entity_length (Optional[int]): The maximum length of the encoded entity inputs. Defaults to None.
            stride (int): The stride to use when truncating the inputs. Defaults to 0.
            is_split_into_words (Optional[bool]): Whether the inputs are already split into words. Defaults to False.
            pad_to_multiple_of (Optional[int]): Pad the inputs to a multiple of this value. Defaults to None.
            return_tensors (Optional[Union[str, TensorType]]): The type of tensor to return. Defaults to None.
            return_token_type_ids (Optional[bool]): Whether to return token type ids. Defaults to None.
            return_attention_mask (Optional[bool]): Whether to return attention masks. Defaults to None.
            return_overflowing_tokens (bool): Whether to return overflowing tokens. Defaults to False.
            return_special_tokens_mask (bool): Whether to return special tokens masks. Defaults to False.
            return_offsets_mapping (bool): Whether to return character offsets mapping. Defaults to False.
            return_length (bool): Whether to return the lengths of the encoded inputs. Defaults to False.
            verbose (bool): Whether to print information about the encoding process. Defaults to True.

        Returns:
            BatchEncoding: The encoded batch inputs.

        Raises:
            NotImplementedError: If return_offsets_mapping is used with a tokenizer that does not support it.
            NotImplementedError: If is_split_into_words is used with this tokenizer.

        """
        if return_offsets_mapping:
            raise NotImplementedError(
                "return_offset_mapping is not available when using Python tokenizers. "
                "To use this feature, change your tokenizer to one deriving from "
                "transformers.PreTrainedTokenizerFast."
            )

        if is_split_into_words:
            raise NotImplementedError("is_split_into_words is not supported in this tokenizer.")

        # input_ids is a list of tuples (one for each example in the batch)
        input_ids = []
        entity_ids = []
        entity_token_spans = []
        for index, text_or_text_pair in enumerate(batch_text_or_text_pairs):
            if not isinstance(text_or_text_pair, (list, tuple)):
                text, text_pair = text_or_text_pair, None
            else:
                text, text_pair = text_or_text_pair

            entities, entities_pair = None, None
            if batch_entities_or_entities_pairs is not None:
                entities_or_entities_pairs = batch_entities_or_entities_pairs[index]
                if entities_or_entities_pairs:
                    if isinstance(entities_or_entities_pairs[0], str):
                        entities, entities_pair = entities_or_entities_pairs, None
                    else:
                        entities, entities_pair = entities_or_entities_pairs

            entity_spans, entity_spans_pair = None, None
            if batch_entity_spans_or_entity_spans_pairs is not None:
                entity_spans_or_entity_spans_pairs = batch_entity_spans_or_entity_spans_pairs[index]
                if len(entity_spans_or_entity_spans_pairs) > 0 and isinstance(
                    entity_spans_or_entity_spans_pairs[0], list
                ):
                    entity_spans, entity_spans_pair = entity_spans_or_entity_spans_pairs
                else:
                    entity_spans, entity_spans_pair = entity_spans_or_entity_spans_pairs, None

            (
                first_ids,
                second_ids,
                first_entity_ids,
                second_entity_ids,
                first_entity_token_spans,
                second_entity_token_spans,
            ) = self._create_input_sequence(
                text=text,
                text_pair=text_pair,
                entities=entities,
                entities_pair=entities_pair,
                entity_spans=entity_spans,
                entity_spans_pair=entity_spans_pair,
                **kwargs,
            )
            input_ids.append((first_ids, second_ids))
            entity_ids.append((first_entity_ids, second_entity_ids))
            entity_token_spans.append((first_entity_token_spans, second_entity_token_spans))

        batch_outputs = self._batch_prepare_for_model(
            input_ids,
            batch_entity_ids_pairs=entity_ids,
            batch_entity_token_spans_pairs=entity_token_spans,
            add_special_tokens=add_special_tokens,
            padding_strategy=padding_strategy,
            truncation_strategy=truncation_strategy,
            max_length=max_length,
            max_entity_length=max_entity_length,
            stride=stride,
            pad_to_multiple_of=pad_to_multiple_of,
            return_attention_mask=return_attention_mask,
            return_token_type_ids=return_token_type_ids,
            return_overflowing_tokens=return_overflowing_tokens,
            return_special_tokens_mask=return_special_tokens_mask,
            return_length=return_length,
            return_tensors=return_tensors,
            verbose=verbose,
        )

        return BatchEncoding(batch_outputs)

    def _check_entity_input_format(self, entities: Optional[EntityInput], entity_spans: Optional[EntitySpanInput]):
        """
        This method '_check_entity_input_format' in the class 'LukeTokenizer' validates the input format for entities and entity spans.

        Args:
            self: The instance of the class.
            entities (Optional[EntityInput]): A list of entity names. If specified, it should be given as a list of entity names.
            entity_spans (Optional[EntitySpanInput]): A list of tuples containing the start and end character indices.
                If specified, it should be given as a list of tuples containing the start and end character indices.

        Returns:
            None.

        Raises:
            ValueError:
                - If 'entity_spans' is not given as a list.
                - If 'entity_spans' is given as a list, but the first element is not a tuple containing
                the start and end character indices.
                - If 'entities' is specified but not given as a list.
                - If 'entities' is given as a list, but the first element is not a string.
                - If the length of 'entities' is not equal to the length of 'entity_spans' when both are specified.
        """
        if not isinstance(entity_spans, list):
            raise ValueError("entity_spans should be given as a list")
        if len(entity_spans) > 0 and not isinstance(entity_spans[0], tuple):
            raise ValueError(
                "entity_spans should be given as a list of tuples containing the start and end character indices"
            )

        if entities is not None:
            if not isinstance(entities, list):
                raise ValueError("If you specify entities, they should be given as a list")

            if len(entities) > 0 and not isinstance(entities[0], str):
                raise ValueError("If you specify entities, they should be given as a list of entity names")

            if len(entities) != len(entity_spans):
                raise ValueError("If you specify entities, entities and entity_spans must be the same length")

    def _create_input_sequence(
        self,
        text: TextInput,
        text_pair: Optional[TextInput] = None,
        entities: Optional[EntityInput] = None,
        entities_pair: Optional[EntityInput] = None,
        entity_spans: Optional[EntitySpanInput] = None,
        entity_spans_pair: Optional[EntitySpanInput] = None,
        **kwargs,
    ) -> Tuple[list, list, list, list, list, list]:
        """
        Create input sequences for the LukeTokenizer.

        Args:
            self (LukeTokenizer): An instance of the LukeTokenizer class.
            text (TextInput): The main input text to be tokenized.
            text_pair (Optional[TextInput]): An optional pair of input text to be tokenized. Default is None.
            entities (Optional[EntityInput]): An optional list of entities in the main input text. Default is None.
            entities_pair (Optional[EntityInput]): An optional list of entities in the pair input text. Default is None.
            entity_spans (Optional[EntitySpanInput]): An optional list of tuples representing the start and end character
                indices of the entities in the main input text. Default is None.
            entity_spans_pair (Optional[EntitySpanInput]): An optional list of tuples representing the start and end
                character indices of the entities in the pair input text. Default is None.
            **kwargs: Additional keyword arguments.

        Returns:
            Tuple[list, list, list, list, list, list]:
                A tuple containing six lists:

                - first_ids: A list of token IDs for the main input text.
                - second_ids: A list of token IDs for the pair input text.
                - first_entity_ids: A list of entity IDs for the entities in the main input text.
                - second_entity_ids: A list of entity IDs for the entities in the pair input text.
                - first_entity_token_spans: A list of token spans for the entities in the main input text.
                - second_entity_token_spans: A list of token spans for the entities in the pair input text.

        Raises:
            ValueError: If the task is not supported or if the entity spans are not in the correct format.
        """
        def get_input_ids(text):
            tokens = self.tokenize(text, **kwargs)
            return self.convert_tokens_to_ids(tokens)

        def get_input_ids_and_entity_token_spans(text, entity_spans):
            if entity_spans is None:
                return get_input_ids(text), None

            cur = 0
            input_ids = []
            entity_token_spans = [None] * len(entity_spans)

            split_char_positions = sorted(frozenset(itertools.chain(*entity_spans)))
            char_pos2token_pos = {}

            for split_char_position in split_char_positions:
                orig_split_char_position = split_char_position
                if (
                    split_char_position > 0 and text[split_char_position - 1] == " "
                ):  # whitespace should be prepended to the following token
                    split_char_position -= 1
                if cur != split_char_position:
                    input_ids += get_input_ids(text[cur:split_char_position])
                    cur = split_char_position
                char_pos2token_pos[orig_split_char_position] = len(input_ids)

            input_ids += get_input_ids(text[cur:])

            entity_token_spans = [
                (char_pos2token_pos[char_start], char_pos2token_pos[char_end]) for char_start, char_end in entity_spans
            ]

            return input_ids, entity_token_spans

        first_ids, second_ids = None, None
        first_entity_ids, second_entity_ids = None, None
        first_entity_token_spans, second_entity_token_spans = None, None

        if self.task is None:
            if entity_spans is None:
                first_ids = get_input_ids(text)
            else:
                self._check_entity_input_format(entities, entity_spans)

                first_ids, first_entity_token_spans = get_input_ids_and_entity_token_spans(text, entity_spans)
                if entities is None:
                    first_entity_ids = [self.entity_mask_token_id] * len(entity_spans)
                else:
                    first_entity_ids = [self.entity_vocab.get(entity, self.entity_unk_token_id) for entity in entities]

            if text_pair is not None:
                if entity_spans_pair is None:
                    second_ids = get_input_ids(text_pair)
                else:
                    self._check_entity_input_format(entities_pair, entity_spans_pair)

                    second_ids, second_entity_token_spans = get_input_ids_and_entity_token_spans(
                        text_pair, entity_spans_pair
                    )
                    if entities_pair is None:
                        second_entity_ids = [self.entity_mask_token_id] * len(entity_spans_pair)
                    else:
                        second_entity_ids = [
                            self.entity_vocab.get(entity, self.entity_unk_token_id) for entity in entities_pair
                        ]

        elif self.task == "entity_classification":
            if not (isinstance(entity_spans, list) and len(entity_spans) == 1 and isinstance(entity_spans[0], tuple)):
                raise ValueError(
                    "Entity spans should be a list containing a single tuple "
                    "containing the start and end character indices of an entity"
                )
            first_entity_ids = [self.entity_mask_token_id]
            first_ids, first_entity_token_spans = get_input_ids_and_entity_token_spans(text, entity_spans)

            # add special tokens to input ids
            entity_token_start, entity_token_end = first_entity_token_spans[0]
            first_ids = (
                first_ids[:entity_token_end] + [self.additional_special_tokens_ids[0]] + first_ids[entity_token_end:]
            )
            first_ids = (
                first_ids[:entity_token_start]
                + [self.additional_special_tokens_ids[0]]
                + first_ids[entity_token_start:]
            )
            first_entity_token_spans = [(entity_token_start, entity_token_end + 2)]

        elif self.task == "entity_pair_classification":
            if not (
                isinstance(entity_spans, list)
                and len(entity_spans) == 2
                and isinstance(entity_spans[0], tuple)
                and isinstance(entity_spans[1], tuple)
            ):
                raise ValueError(
                    "Entity spans should be provided as a list of two tuples, "
                    "each tuple containing the start and end character indices of an entity"
                )

            first_entity_ids = [self.entity_mask_token_id, self.entity_mask2_token_id]
            first_ids, first_entity_token_spans = get_input_ids_and_entity_token_spans(text, entity_spans)

            head_token_span, tail_token_span = first_entity_token_spans
            token_span_with_special_token_ids = [
                (head_token_span, self.additional_special_tokens_ids[0]),
                (tail_token_span, self.additional_special_tokens_ids[1]),
            ]
            if head_token_span[0] < tail_token_span[0]:
                first_entity_token_spans[0] = (head_token_span[0], head_token_span[1] + 2)
                first_entity_token_spans[1] = (tail_token_span[0] + 2, tail_token_span[1] + 4)
                token_span_with_special_token_ids = reversed(token_span_with_special_token_ids)
            else:
                first_entity_token_spans[0] = (head_token_span[0] + 2, head_token_span[1] + 4)
                first_entity_token_spans[1] = (tail_token_span[0], tail_token_span[1] + 2)

            for (entity_token_start, entity_token_end), special_token_id in token_span_with_special_token_ids:
                first_ids = first_ids[:entity_token_end] + [special_token_id] + first_ids[entity_token_end:]
                first_ids = first_ids[:entity_token_start] + [special_token_id] + first_ids[entity_token_start:]

        elif self.task == "entity_span_classification":
            if not (isinstance(entity_spans, list) and len(entity_spans) > 0 and isinstance(entity_spans[0], tuple)):
                raise ValueError(
                    "Entity spans should be provided as a list of tuples, "
                    "each tuple containing the start and end character indices of an entity"
                )

            first_ids, first_entity_token_spans = get_input_ids_and_entity_token_spans(text, entity_spans)
            first_entity_ids = [self.entity_mask_token_id] * len(entity_spans)

        else:
            raise ValueError(f"Task {self.task} not supported")

        return (
            first_ids,
            second_ids,
            first_entity_ids,
            second_entity_ids,
            first_entity_token_spans,
            second_entity_token_spans,
        )

    def _batch_prepare_for_model(
        self,
        batch_ids_pairs: List[Tuple[List[int], None]],
        batch_entity_ids_pairs: List[Tuple[Optional[List[int]], Optional[List[int]]]],
        batch_entity_token_spans_pairs: List[Tuple[Optional[List[Tuple[int, int]]], Optional[List[Tuple[int, int]]]]],
        add_special_tokens: bool = True,
        padding_strategy: PaddingStrategy = PaddingStrategy.DO_NOT_PAD,
        truncation_strategy: TruncationStrategy = TruncationStrategy.DO_NOT_TRUNCATE,
        max_length: Optional[int] = None,
        max_entity_length: Optional[int] = None,
        stride: int = 0,
        pad_to_multiple_of: Optional[int] = None,
        return_tensors: Optional[str] = None,
        return_token_type_ids: Optional[bool] = None,
        return_attention_mask: Optional[bool] = None,
        return_overflowing_tokens: bool = False,
        return_special_tokens_mask: bool = False,
        return_length: bool = False,
        verbose: bool = True,
    ) -> BatchEncoding:
        """
        Prepares a sequence of input id, or a pair of sequences of inputs ids so that it can be used by the model. It
        adds special tokens, truncates sequences if overflowing while taking into account the special tokens and
        manages a moving window (with user defined stride) for overflowing tokens


        Args:
            batch_ids_pairs: list of tokenized input ids or input ids pairs
            batch_entity_ids_pairs: list of entity ids or entity ids pairs
            batch_entity_token_spans_pairs: list of entity spans or entity spans pairs
            max_entity_length: The maximum length of the entity sequence.
        """
        batch_outputs = {}
        for input_ids, entity_ids, entity_token_span_pairs in zip(
            batch_ids_pairs, batch_entity_ids_pairs, batch_entity_token_spans_pairs
        ):
            first_ids, second_ids = input_ids
            first_entity_ids, second_entity_ids = entity_ids
            first_entity_token_spans, second_entity_token_spans = entity_token_span_pairs
            outputs = self.prepare_for_model(
                first_ids,
                second_ids,
                entity_ids=first_entity_ids,
                pair_entity_ids=second_entity_ids,
                entity_token_spans=first_entity_token_spans,
                pair_entity_token_spans=second_entity_token_spans,
                add_special_tokens=add_special_tokens,
                padding=PaddingStrategy.DO_NOT_PAD.value,  # we pad in batch afterward
                truncation=truncation_strategy.value,
                max_length=max_length,
                max_entity_length=max_entity_length,
                stride=stride,
                pad_to_multiple_of=None,  # we pad in batch afterward
                return_attention_mask=False,  # we pad in batch afterward
                return_token_type_ids=return_token_type_ids,
                return_overflowing_tokens=return_overflowing_tokens,
                return_special_tokens_mask=return_special_tokens_mask,
                return_length=return_length,
                return_tensors=None,  # We convert the whole batch to tensors at the end
                prepend_batch_axis=False,
                verbose=verbose,
            )

            for key, value in outputs.items():
                if key not in batch_outputs:
                    batch_outputs[key] = []
                batch_outputs[key].append(value)

        batch_outputs = self.pad(
            batch_outputs,
            padding=padding_strategy.value,
            max_length=max_length,
            pad_to_multiple_of=pad_to_multiple_of,
            return_attention_mask=return_attention_mask,
        )

        batch_outputs = BatchEncoding(batch_outputs, tensor_type=return_tensors)

        return batch_outputs

    def prepare_for_model(
        self,
        ids: List[int],
        pair_ids: Optional[List[int]] = None,
        entity_ids: Optional[List[int]] = None,
        pair_entity_ids: Optional[List[int]] = None,
        entity_token_spans: Optional[List[Tuple[int, int]]] = None,
        pair_entity_token_spans: Optional[List[Tuple[int, int]]] = None,
        add_special_tokens: bool = True,
        padding: Union[bool, str, PaddingStrategy] = False,
        truncation: Union[bool, str, TruncationStrategy] = None,
        max_length: Optional[int] = None,
        max_entity_length: Optional[int] = None,
        stride: int = 0,
        pad_to_multiple_of: Optional[int] = None,
        return_tensors: Optional[Union[str, TensorType]] = None,
        return_token_type_ids: Optional[bool] = None,
        return_attention_mask: Optional[bool] = None,
        return_overflowing_tokens: bool = False,
        return_special_tokens_mask: bool = False,
        return_offsets_mapping: bool = False,
        return_length: bool = False,
        verbose: bool = True,
        prepend_batch_axis: bool = False,
        **kwargs,
    ) -> BatchEncoding:
        """
        Prepares a sequence of input id, entity id and entity span, or a pair of sequences of inputs ids, entity ids,
        entity spans so that it can be used by the model. It adds special tokens, truncates sequences if overflowing
        while taking into account the special tokens and manages a moving window (with user defined stride) for
        overflowing tokens. Please Note, for *pair_ids* different than `None` and *truncation_strategy = longest_first*
        or `True`, it is not possible to return overflowing tokens. Such a combination of arguments will raise an
        error.

        Args:
            ids (`List[int]`):
                Tokenized input ids of the first sequence.
            pair_ids (`List[int]`, *optional*):
                Tokenized input ids of the second sequence.
            entity_ids (`List[int]`, *optional*):
                Entity ids of the first sequence.
            pair_entity_ids (`List[int]`, *optional*):
                Entity ids of the second sequence.
            entity_token_spans (`List[Tuple[int, int]]`, *optional*):
                Entity spans of the first sequence.
            pair_entity_token_spans (`List[Tuple[int, int]]`, *optional*):
                Entity spans of the second sequence.
            max_entity_length (`int`, *optional*):
                The maximum length of the entity sequence.
        """
        # Backward compatibility for 'truncation_strategy', 'pad_to_max_length'
        padding_strategy, truncation_strategy, max_length, kwargs = self._get_padding_truncation_strategies(
            padding=padding,
            truncation=truncation,
            max_length=max_length,
            pad_to_multiple_of=pad_to_multiple_of,
            verbose=verbose,
            **kwargs,
        )

        # Compute lengths
        pair = bool(pair_ids is not None)
        len_ids = len(ids)
        len_pair_ids = len(pair_ids) if pair else 0

        if return_token_type_ids and not add_special_tokens:
            raise ValueError(
                "Asking to return token_type_ids while setting add_special_tokens to False "
                "results in an undefined behavior. Please set add_special_tokens to True or "
                "set return_token_type_ids to None."
            )
        if (
            return_overflowing_tokens
            and truncation_strategy == TruncationStrategy.LONGEST_FIRST
            and pair_ids is not None
        ):
            raise ValueError(
                "Not possible to return overflowing tokens for pair of sequences with the "
                "`longest_first`. Please select another truncation strategy than `longest_first`, "
                "for instance `only_second` or `only_first`."
            )

        # Load from model defaults
        if return_token_type_ids is None:
            return_token_type_ids = "token_type_ids" in self.model_input_names
        if return_attention_mask is None:
            return_attention_mask = "attention_mask" in self.model_input_names

        encoded_inputs = {}

        # Compute the total size of the returned word encodings
        total_len = len_ids + len_pair_ids + (self.num_special_tokens_to_add(pair=pair) if add_special_tokens else 0)

        # Truncation: Handle max sequence length and max_entity_length
        overflowing_tokens = []
        if truncation_strategy != TruncationStrategy.DO_NOT_TRUNCATE and max_length and total_len > max_length:
            # truncate words up to max_length
            ids, pair_ids, overflowing_tokens = self.truncate_sequences(
                ids,
                pair_ids=pair_ids,
                num_tokens_to_remove=total_len - max_length,
                truncation_strategy=truncation_strategy,
                stride=stride,
            )

        if return_overflowing_tokens:
            encoded_inputs["overflowing_tokens"] = overflowing_tokens
            encoded_inputs["num_truncated_tokens"] = total_len - max_length

        # Add special tokens
        if add_special_tokens:
            sequence = self.build_inputs_with_special_tokens(ids, pair_ids)
            token_type_ids = self.create_token_type_ids_from_sequences(ids, pair_ids)
            entity_token_offset = 1  # 1 * <s> token
            pair_entity_token_offset = len(ids) + 3  # 1 * <s> token & 2 * <sep> tokens
        else:
            sequence = ids + pair_ids if pair else ids
            token_type_ids = [0] * len(ids) + ([0] * len(pair_ids) if pair else [])
            entity_token_offset = 0
            pair_entity_token_offset = len(ids)

        # Build output dictionary
        encoded_inputs["input_ids"] = sequence
        if return_token_type_ids:
            encoded_inputs["token_type_ids"] = token_type_ids
        if return_special_tokens_mask:
            if add_special_tokens:
                encoded_inputs["special_tokens_mask"] = self.get_special_tokens_mask(ids, pair_ids)
            else:
                encoded_inputs["special_tokens_mask"] = [0] * len(sequence)

        # Set max entity length
        if not max_entity_length:
            max_entity_length = self.max_entity_length

        if entity_ids is not None:
            total_entity_len = 0
            num_invalid_entities = 0
            valid_entity_ids = [ent_id for ent_id, span in zip(entity_ids, entity_token_spans) if span[1] <= len(ids)]
            valid_entity_token_spans = [span for span in entity_token_spans if span[1] <= len(ids)]

            total_entity_len += len(valid_entity_ids)
            num_invalid_entities += len(entity_ids) - len(valid_entity_ids)

            valid_pair_entity_ids, valid_pair_entity_token_spans = None, None
            if pair_entity_ids is not None:
                valid_pair_entity_ids = [
                    ent_id
                    for ent_id, span in zip(pair_entity_ids, pair_entity_token_spans)
                    if span[1] <= len(pair_ids)
                ]
                valid_pair_entity_token_spans = [span for span in pair_entity_token_spans if span[1] <= len(pair_ids)]
                total_entity_len += len(valid_pair_entity_ids)
                num_invalid_entities += len(pair_entity_ids) - len(valid_pair_entity_ids)

            if num_invalid_entities != 0:
                logger.warning(
                    f"{num_invalid_entities} entities are ignored because their entity spans are invalid due to the"
                    " truncation of input tokens"
                )

            if truncation_strategy != TruncationStrategy.DO_NOT_TRUNCATE and total_entity_len > max_entity_length:
                # truncate entities up to max_entity_length
                valid_entity_ids, valid_pair_entity_ids, overflowing_entities = self.truncate_sequences(
                    valid_entity_ids,
                    pair_ids=valid_pair_entity_ids,
                    num_tokens_to_remove=total_entity_len - max_entity_length,
                    truncation_strategy=truncation_strategy,
                    stride=stride,
                )
                valid_entity_token_spans = valid_entity_token_spans[: len(valid_entity_ids)]
                if valid_pair_entity_token_spans is not None:
                    valid_pair_entity_token_spans = valid_pair_entity_token_spans[: len(valid_pair_entity_ids)]

            if return_overflowing_tokens:
                encoded_inputs["overflowing_entities"] = overflowing_entities
                encoded_inputs["num_truncated_entities"] = total_entity_len - max_entity_length

            final_entity_ids = valid_entity_ids + valid_pair_entity_ids if valid_pair_entity_ids else valid_entity_ids
            encoded_inputs["entity_ids"] = list(final_entity_ids)
            entity_position_ids = []
            entity_start_positions = []
            entity_end_positions = []
            for token_spans, offset in (
                (valid_entity_token_spans, entity_token_offset),
                (valid_pair_entity_token_spans, pair_entity_token_offset),
            ):
                if token_spans is not None:
                    for start, end in token_spans:
                        start += offset
                        end += offset
                        position_ids = list(range(start, end))[: self.max_mention_length]
                        position_ids += [-1] * (self.max_mention_length - end + start)
                        entity_position_ids.append(position_ids)
                        entity_start_positions.append(start)
                        entity_end_positions.append(end - 1)

            encoded_inputs["entity_position_ids"] = entity_position_ids
            if self.task == "entity_span_classification":
                encoded_inputs["entity_start_positions"] = entity_start_positions
                encoded_inputs["entity_end_positions"] = entity_end_positions

            if return_token_type_ids:
                encoded_inputs["entity_token_type_ids"] = [0] * len(encoded_inputs["entity_ids"])

        # Check lengths
        self._eventual_warn_about_too_long_sequence(encoded_inputs["input_ids"], max_length, verbose)

        # Padding
        if padding_strategy != PaddingStrategy.DO_NOT_PAD or return_attention_mask:
            encoded_inputs = self.pad(
                encoded_inputs,
                max_length=max_length,
                max_entity_length=max_entity_length,
                padding=padding_strategy.value,
                pad_to_multiple_of=pad_to_multiple_of,
                return_attention_mask=return_attention_mask,
            )

        if return_length:
            encoded_inputs["length"] = len(encoded_inputs["input_ids"])

        batch_outputs = BatchEncoding(
            encoded_inputs, tensor_type=return_tensors, prepend_batch_axis=prepend_batch_axis
        )

        return batch_outputs

    def pad(
        self,
        encoded_inputs: Union[
            BatchEncoding,
            List[BatchEncoding],
            Dict[str, EncodedInput],
            Dict[str, List[EncodedInput]],
            List[Dict[str, EncodedInput]],
        ],
        padding: Union[bool, str, PaddingStrategy] = True,
        max_length: Optional[int] = None,
        max_entity_length: Optional[int] = None,
        pad_to_multiple_of: Optional[int] = None,
        return_attention_mask: Optional[bool] = None,
        return_tensors: Optional[Union[str, TensorType]] = None,
        verbose: bool = True,
    ) -> BatchEncoding:
        """
        Pad a single encoded input or a batch of encoded inputs up to predefined length or to the max sequence length
        in the batch. Padding side (left/right) padding token ids are defined at the tokenizer level (with
        `self.padding_side`, `self.pad_token_id` and `self.pad_token_type_id`) .. note:: If the `encoded_inputs` passed
        are dictionary of numpy arrays, PyTorch tensors or TensorFlow tensors, the result will use the same type unless
        you provide a different tensor type with `return_tensors`. In the case of PyTorch tensors, you will lose the
        specific device of your tensors however.

        Args:
            encoded_inputs ([`BatchEncoding`], list of [`BatchEncoding`], `Dict[str, List[int]]`, `Dict[str, List[List[int]]` or `List[Dict[str, List[int]]]`):
                Tokenized inputs. Can represent one input ([`BatchEncoding`] or `Dict[str, List[int]]`) or a batch of
                tokenized inputs (list of [`BatchEncoding`], *Dict[str, List[List[int]]]* or *List[Dict[str,
                List[int]]]*) so you can use this method during preprocessing as well as in a PyTorch Dataloader
                collate function. Instead of `List[int]` you can have tensors (numpy arrays, PyTorch tensors or
                TensorFlow tensors), see the note above for the return type.
            padding (`bool`, `str` or [`~utils.PaddingStrategy`], *optional*, defaults to `True`):
                 Select a strategy to pad the returned sequences (according to the model's padding side and padding
                 index) among:

                - `True` or `'longest'`: Pad to the longest sequence in the batch (or no padding if only a single
                sequence if provided).
                - `'max_length'`: Pad to a maximum length specified with the argument `max_length` or to the maximum
                acceptable input length for the model if that argument is not provided.
                - `False` or `'do_not_pad'` (default): No padding (i.e., can output a batch with sequences of different
                lengths).
            max_length (`int`, *optional*):
                Maximum length of the returned list and optionally padding length (see above).
            max_entity_length (`int`, *optional*):
                The maximum length of the entity sequence.
            pad_to_multiple_of (`int`, *optional*):
                If set will pad the sequence to a multiple of the provided value. This is especially useful to enable
                the use of Tensor Cores on NVIDIA hardware with compute capability `>= 7.5` (Volta).
            return_attention_mask (`bool`, *optional*):
                Whether to return the attention mask. If left to the default, will return the attention mask according
                to the specific tokenizer's default, defined by the `return_outputs` attribute. [What are attention
                masks?](../glossary#attention-mask)
            return_tensors (`str` or [`~utils.TensorType`], *optional*):
                If set, will return tensors instead of list of python integers. Acceptable values are:

                - `'tf'`: Return TensorFlow `tf.constant` objects.
                - `'pt'`: Return PyTorch `torch.Tensor` objects.
                - `'np'`: Return Numpy `np.ndarray` objects.
            verbose (`bool`, *optional*, defaults to `True`):
                Whether or not to print more information and warnings.
        """
        # If we have a list of dicts, let's convert it in a dict of lists
        # We do this to allow using this method as a collate_fn function in PyTorch Dataloader
        if isinstance(encoded_inputs, (list, tuple)) and isinstance(encoded_inputs[0], Mapping):
            encoded_inputs = {key: [example[key] for example in encoded_inputs] for key in encoded_inputs[0].keys()}

        # The model's main input name, usually `input_ids`, has be passed for padding
        if self.model_input_names[0] not in encoded_inputs:
            raise ValueError(
                "You should supply an encoding or a list of encodings to this method "
                f"that includes {self.model_input_names[0]}, but you provided {list(encoded_inputs.keys())}"
            )

        required_input = encoded_inputs[self.model_input_names[0]]

        if not required_input:
            if return_attention_mask:
                encoded_inputs["attention_mask"] = []
            return encoded_inputs

        # If we have PyTorch/TF/NumPy tensors/arrays as inputs, we cast them as python objects
        # and rebuild them afterwards if no return_tensors is specified
        # Note that we lose the specific device the tensor may be on for PyTorch

        first_element = required_input[0]
        if isinstance(first_element, (list, tuple)):
            # first_element might be an empty list/tuple in some edge cases so we grab the first non empty element.
            index = 0
            while len(required_input[index]) == 0:
                index += 1
            if index < len(required_input):
                first_element = required_input[index][0]
        # At this state, if `first_element` is still a list/tuple, it's an empty one so there is nothing to do.
        if not isinstance(first_element, (int, list, tuple)):
            if is_mindspore_tensor(first_element):
                return_tensors = "ms" if return_tensors is None else return_tensors
            elif isinstance(first_element, np.ndarray):
                return_tensors = "np" if return_tensors is None else return_tensors
            else:
                raise ValueError(
                    f"type of {first_element} unknown: {type(first_element)}. "
                    "Should be one of a python, numpy, pytorch or tensorflow object."
                )

            for key, value in encoded_inputs.items():
                encoded_inputs[key] = to_py_obj(value)

        # Convert padding_strategy in PaddingStrategy
        padding_strategy, _, max_length, _ = self._get_padding_truncation_strategies(
            padding=padding, max_length=max_length, verbose=verbose
        )

        if max_entity_length is None:
            max_entity_length = self.max_entity_length

        required_input = encoded_inputs[self.model_input_names[0]]
        if required_input and not isinstance(required_input[0], (list, tuple)):
            encoded_inputs = self._pad(
                encoded_inputs,
                max_length=max_length,
                max_entity_length=max_entity_length,
                padding_strategy=padding_strategy,
                pad_to_multiple_of=pad_to_multiple_of,
                return_attention_mask=return_attention_mask,
            )
            return BatchEncoding(encoded_inputs, tensor_type=return_tensors)

        batch_size = len(required_input)
        if any(len(v) != batch_size for v in encoded_inputs.values()):
            raise ValueError("Some items in the output dictionary have a different batch size than others.")

        if padding_strategy == PaddingStrategy.LONGEST:
            max_length = max(len(inputs) for inputs in required_input)
            max_entity_length = (
                max(len(inputs) for inputs in encoded_inputs["entity_ids"]) if "entity_ids" in encoded_inputs else 0
            )
            padding_strategy = PaddingStrategy.MAX_LENGTH

        batch_outputs = {}
        for i in range(batch_size):
            inputs = {k: v[i] for k, v in encoded_inputs.items()}
            outputs = self._pad(
                inputs,
                max_length=max_length,
                max_entity_length=max_entity_length,
                padding_strategy=padding_strategy,
                pad_to_multiple_of=pad_to_multiple_of,
                return_attention_mask=return_attention_mask,
            )

            for key, value in outputs.items():
                if key not in batch_outputs:
                    batch_outputs[key] = []
                batch_outputs[key].append(value)

        return BatchEncoding(batch_outputs, tensor_type=return_tensors)

    def _pad(
        self,
        encoded_inputs: Union[Dict[str, EncodedInput], BatchEncoding],
        max_length: Optional[int] = None,
        max_entity_length: Optional[int] = None,
        padding_strategy: PaddingStrategy = PaddingStrategy.DO_NOT_PAD,
        pad_to_multiple_of: Optional[int] = None,
        return_attention_mask: Optional[bool] = None,
    ) -> dict:
        """
        Pad encoded inputs (on left/right and up to predefined length or max length in the batch)


        Args:
            encoded_inputs:
                Dictionary of tokenized inputs (`List[int]`) or batch of tokenized inputs (`List[List[int]]`).
            max_length: maximum length of the returned list and optionally padding length (see below).
                Will truncate by taking into account the special tokens.
            max_entity_length: The maximum length of the entity sequence.
            padding_strategy:
                PaddingStrategy to use for padding.

                - PaddingStrategy.LONGEST Pad to the longest sequence in the batch
                - PaddingStrategy.MAX_LENGTH: Pad to the max length (default)
                - PaddingStrategy.DO_NOT_PAD: Do not pad

                The tokenizer padding sides are defined in self.padding_side:

                - 'left': pads on the left of the sequences
                - 'right': pads on the right of the sequences
            pad_to_multiple_of: (optional) Integer if set will pad the sequence to a multiple of the provided value.
                This is especially useful to enable the use of Tensor Core on NVIDIA hardware with compute capability
                `>= 7.5` (Volta).
            return_attention_mask:
                (optional) Set to False to avoid returning attention mask (default: set to model specifics)
        """
        entities_provided = bool("entity_ids" in encoded_inputs)

        # Load from model defaults
        if return_attention_mask is None:
            return_attention_mask = "attention_mask" in self.model_input_names

        if padding_strategy == PaddingStrategy.LONGEST:
            max_length = len(encoded_inputs["input_ids"])
            if entities_provided:
                max_entity_length = len(encoded_inputs["entity_ids"])

        if max_length is not None and pad_to_multiple_of is not None and (max_length % pad_to_multiple_of != 0):
            max_length = ((max_length // pad_to_multiple_of) + 1) * pad_to_multiple_of

        if (
            entities_provided
            and max_entity_length is not None
            and pad_to_multiple_of is not None
            and (max_entity_length % pad_to_multiple_of != 0)
        ):
            max_entity_length = ((max_entity_length // pad_to_multiple_of) + 1) * pad_to_multiple_of

        needs_to_be_padded = padding_strategy != PaddingStrategy.DO_NOT_PAD and (
            len(encoded_inputs["input_ids"]) != max_length
            or (entities_provided and len(encoded_inputs["entity_ids"]) != max_entity_length)
        )

        # Initialize attention mask if not present.
        if return_attention_mask and "attention_mask" not in encoded_inputs:
            encoded_inputs["attention_mask"] = [1] * len(encoded_inputs["input_ids"])
        if entities_provided and return_attention_mask and "entity_attention_mask" not in encoded_inputs:
            encoded_inputs["entity_attention_mask"] = [1] * len(encoded_inputs["entity_ids"])

        if needs_to_be_padded:
            difference = max_length - len(encoded_inputs["input_ids"])
            if entities_provided:
                entity_difference = max_entity_length - len(encoded_inputs["entity_ids"])
            if self.padding_side == "right":
                if return_attention_mask:
                    encoded_inputs["attention_mask"] = encoded_inputs["attention_mask"] + [0] * difference
                    if entities_provided:
                        encoded_inputs["entity_attention_mask"] = (
                            encoded_inputs["entity_attention_mask"] + [0] * entity_difference
                        )
                if "token_type_ids" in encoded_inputs:
                    encoded_inputs["token_type_ids"] = encoded_inputs["token_type_ids"] + [0] * difference
                    if entities_provided:
                        encoded_inputs["entity_token_type_ids"] = (
                            encoded_inputs["entity_token_type_ids"] + [0] * entity_difference
                        )
                if "special_tokens_mask" in encoded_inputs:
                    encoded_inputs["special_tokens_mask"] = encoded_inputs["special_tokens_mask"] + [1] * difference
                encoded_inputs["input_ids"] = encoded_inputs["input_ids"] + [self.pad_token_id] * difference
                if entities_provided:
                    encoded_inputs["entity_ids"] = (
                        encoded_inputs["entity_ids"] + [self.entity_pad_token_id] * entity_difference
                    )
                    encoded_inputs["entity_position_ids"] = (
                        encoded_inputs["entity_position_ids"] + [[-1] * self.max_mention_length] * entity_difference
                    )
                    if self.task == "entity_span_classification":
                        encoded_inputs["entity_start_positions"] = (
                            encoded_inputs["entity_start_positions"] + [0] * entity_difference
                        )
                        encoded_inputs["entity_end_positions"] = (
                            encoded_inputs["entity_end_positions"] + [0] * entity_difference
                        )

            elif self.padding_side == "left":
                if return_attention_mask:
                    encoded_inputs["attention_mask"] = [0] * difference + encoded_inputs["attention_mask"]
                    if entities_provided:
                        encoded_inputs["entity_attention_mask"] = [0] * entity_difference + encoded_inputs[
                            "entity_attention_mask"
                        ]
                if "token_type_ids" in encoded_inputs:
                    encoded_inputs["token_type_ids"] = [0] * difference + encoded_inputs["token_type_ids"]
                    if entities_provided:
                        encoded_inputs["entity_token_type_ids"] = [0] * entity_difference + encoded_inputs[
                            "entity_token_type_ids"
                        ]
                if "special_tokens_mask" in encoded_inputs:
                    encoded_inputs["special_tokens_mask"] = [1] * difference + encoded_inputs["special_tokens_mask"]
                encoded_inputs["input_ids"] = [self.pad_token_id] * difference + encoded_inputs["input_ids"]
                if entities_provided:
                    encoded_inputs["entity_ids"] = [self.entity_pad_token_id] * entity_difference + encoded_inputs[
                        "entity_ids"
                    ]
                    encoded_inputs["entity_position_ids"] = [
                        [-1] * self.max_mention_length
                    ] * entity_difference + encoded_inputs["entity_position_ids"]
                    if self.task == "entity_span_classification":
                        encoded_inputs["entity_start_positions"] = [0] * entity_difference + encoded_inputs[
                            "entity_start_positions"
                        ]
                        encoded_inputs["entity_end_positions"] = [0] * entity_difference + encoded_inputs[
                            "entity_end_positions"
                        ]
            else:
                raise ValueError("Invalid padding strategy:" + str(self.padding_side))

        return encoded_inputs

    def save_vocabulary(self, save_directory: str, filename_prefix: Optional[str] = None) -> Tuple[str]:
        """
        Save the vocabulary to specified directory with an optional filename prefix.

        Args:
            self: Instance of LukeTokenizer class.
            save_directory (str): The directory path where the vocabulary files will be saved.
            filename_prefix (Optional[str]): An optional prefix to be added to the filename. Default is None.

        Returns:
            Tuple[str]: A tuple containing paths to the saved vocabulary files - vocab_file, merge_file,
                and entity_vocab_file.

        Raises:
            FileNotFoundError: If the specified save_directory does not exist.
            IOError: If there is an issue with reading or writing the vocabulary files.
            ValueError: If the provided filename_prefix is not a string.
            Exception: Any other unexpected error that may occur during the execution of the method.
        """
        if not os.path.isdir(save_directory):
            logger.error(f"Vocabulary path ({save_directory}) should be a directory")
            return
        vocab_file = os.path.join(
            save_directory, (filename_prefix + "-" if filename_prefix else "") + VOCAB_FILES_NAMES["vocab_file"]
        )
        merge_file = os.path.join(
            save_directory, (filename_prefix + "-" if filename_prefix else "") + VOCAB_FILES_NAMES["merges_file"]
        )

        with open(vocab_file, "w", encoding="utf-8") as f:
            f.write(json.dumps(self.encoder, indent=2, sort_keys=True, ensure_ascii=False) + "\n")

        index = 0
        with open(merge_file, "w", encoding="utf-8") as writer:
            writer.write("#version: 0.2\n")
            for bpe_tokens, token_index in sorted(self.bpe_ranks.items(), key=lambda kv: kv[1]):
                if index != token_index:
                    logger.warning(
                        f"Saving vocabulary to {merge_file}: BPE merge indices are not consecutive."
                        " Please check that the tokenizer is not corrupted!"
                    )
                    index = token_index
                writer.write(" ".join(bpe_tokens) + "\n")
                index += 1

        entity_vocab_file = os.path.join(
            save_directory, (filename_prefix + "-" if filename_prefix else "") + VOCAB_FILES_NAMES["entity_vocab_file"]
        )

        with open(entity_vocab_file, "w", encoding="utf-8") as f:
            f.write(json.dumps(self.entity_vocab, indent=2, sort_keys=True, ensure_ascii=False) + "\n")

        return vocab_file, merge_file, entity_vocab_file

mindnlp.transformers.models.luke.tokenization_luke.LukeTokenizer.vocab_size property

Returns the size of the vocabulary.

PARAMETER DESCRIPTION
self

The instance of the LukeTokenizer class.

TYPE: LukeTokenizer

RETURNS DESCRIPTION
int

The number of items in the encoder, representing the size of the vocabulary.

mindnlp.transformers.models.luke.tokenization_luke.LukeTokenizer.__call__(text, text_pair=None, entity_spans=None, entity_spans_pair=None, entities=None, entities_pair=None, add_special_tokens=True, padding=False, truncation=None, max_length=None, max_entity_length=None, stride=0, is_split_into_words=False, pad_to_multiple_of=None, return_tensors=None, return_token_type_ids=None, return_attention_mask=None, return_overflowing_tokens=False, return_special_tokens_mask=False, return_offsets_mapping=False, return_length=False, verbose=True, **kwargs)

Main method to tokenize and prepare for the model one or several sequence(s) or one or several pair(s) of sequences, depending on the task you want to prepare them for.

PARAMETER DESCRIPTION
text

The sequence or batch of sequences to be encoded. Each sequence must be a string. Note that this tokenizer does not support tokenization based on pretokenized strings.

TYPE: `str`, `List[str]`, `List[List[str]]`

text_pair

The sequence or batch of sequences to be encoded. Each sequence must be a string. Note that this tokenizer does not support tokenization based on pretokenized strings.

TYPE: `str`, `List[str]`, `List[List[str]]` DEFAULT: None

entity_spans

The sequence or batch of sequences of entity spans to be encoded. Each sequence consists of tuples each with two integers denoting character-based start and end positions of entities. If you specify "entity_classification" or "entity_pair_classification" as the task argument in the forwardor, the length of each sequence must be 1 or 2, respectively. If you specify entities, the length of each sequence must be equal to the length of each sequence of entities.

TYPE: `List[Tuple[int, int]]`, `List[List[Tuple[int, int]]]`, *optional* DEFAULT: None

entity_spans_pair

The sequence or batch of sequences of entity spans to be encoded. Each sequence consists of tuples each with two integers denoting character-based start and end positions of entities. If you specify the task argument in the forwardor, this argument is ignored. If you specify entities_pair, the length of each sequence must be equal to the length of each sequence of entities_pair.

TYPE: `List[Tuple[int, int]]`, `List[List[Tuple[int, int]]]`, *optional* DEFAULT: None

entities

The sequence or batch of sequences of entities to be encoded. Each sequence consists of strings representing entities, i.e., special entities (e.g., [MASK]) or entity titles of Wikipedia (e.g., Los Angeles). This argument is ignored if you specify the task argument in the forwardor. The length of each sequence must be equal to the length of each sequence of entity_spans. If you specify entity_spans without specifying this argument, the entity sequence or the batch of entity sequences is automatically forwarded by filling it with the [MASK] entity.

TYPE: `List[str]`, `List[List[str]]`, *optional* DEFAULT: None

entities_pair

The sequence or batch of sequences of entities to be encoded. Each sequence consists of strings representing entities, i.e., special entities (e.g., [MASK]) or entity titles of Wikipedia (e.g., Los Angeles). This argument is ignored if you specify the task argument in the forwardor. The length of each sequence must be equal to the length of each sequence of entity_spans_pair. If you specify entity_spans_pair without specifying this argument, the entity sequence or the batch of entity sequences is automatically forwarded by filling it with the [MASK] entity.

TYPE: `List[str]`, `List[List[str]]`, *optional* DEFAULT: None

max_entity_length

The maximum length of entity_ids.

TYPE: `int`, *optional* DEFAULT: None

Source code in mindnlp\transformers\models\luke\tokenization_luke.py 
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
def __call__(
    self,
    text: Union[TextInput, List[TextInput]],
    text_pair: Optional[Union[TextInput, List[TextInput]]] = None,
    entity_spans: Optional[Union[EntitySpanInput, List[EntitySpanInput]]] = None,
    entity_spans_pair: Optional[Union[EntitySpanInput, List[EntitySpanInput]]] = None,
    entities: Optional[Union[EntityInput, List[EntityInput]]] = None,
    entities_pair: Optional[Union[EntityInput, List[EntityInput]]] = None,
    add_special_tokens: bool = True,
    padding: Union[bool, str, PaddingStrategy] = False,
    truncation: Union[bool, str, TruncationStrategy] = None,
    max_length: Optional[int] = None,
    max_entity_length: Optional[int] = None,
    stride: int = 0,
    is_split_into_words: Optional[bool] = False,
    pad_to_multiple_of: Optional[int] = None,
    return_tensors: Optional[Union[str, TensorType]] = None,
    return_token_type_ids: Optional[bool] = None,
    return_attention_mask: Optional[bool] = None,
    return_overflowing_tokens: bool = False,
    return_special_tokens_mask: bool = False,
    return_offsets_mapping: bool = False,
    return_length: bool = False,
    verbose: bool = True,
    **kwargs,
) -> BatchEncoding:
    """
    Main method to tokenize and prepare for the model one or several sequence(s) or one or several pair(s) of
    sequences, depending on the task you want to prepare them for.

    Args:
        text (`str`, `List[str]`, `List[List[str]]`):
            The sequence or batch of sequences to be encoded. Each sequence must be a string. Note that this
            tokenizer does not support tokenization based on pretokenized strings.
        text_pair (`str`, `List[str]`, `List[List[str]]`):
            The sequence or batch of sequences to be encoded. Each sequence must be a string. Note that this
            tokenizer does not support tokenization based on pretokenized strings.
        entity_spans (`List[Tuple[int, int]]`, `List[List[Tuple[int, int]]]`, *optional*):
            The sequence or batch of sequences of entity spans to be encoded. Each sequence consists of tuples each
            with two integers denoting character-based start and end positions of entities. If you specify
            `"entity_classification"` or `"entity_pair_classification"` as the `task` argument in the forwardor,
            the length of each sequence must be 1 or 2, respectively. If you specify `entities`, the length of each
            sequence must be equal to the length of each sequence of `entities`.
        entity_spans_pair (`List[Tuple[int, int]]`, `List[List[Tuple[int, int]]]`, *optional*):
            The sequence or batch of sequences of entity spans to be encoded. Each sequence consists of tuples each
            with two integers denoting character-based start and end positions of entities. If you specify the
            `task` argument in the forwardor, this argument is ignored. If you specify `entities_pair`, the
            length of each sequence must be equal to the length of each sequence of `entities_pair`.
        entities (`List[str]`, `List[List[str]]`, *optional*):
            The sequence or batch of sequences of entities to be encoded. Each sequence consists of strings
            representing entities, i.e., special entities (e.g., [MASK]) or entity titles of Wikipedia (e.g., Los
            Angeles). This argument is ignored if you specify the `task` argument in the forwardor. The length of
            each sequence must be equal to the length of each sequence of `entity_spans`. If you specify
            `entity_spans` without specifying this argument, the entity sequence or the batch of entity sequences
            is automatically forwarded by filling it with the [MASK] entity.
        entities_pair (`List[str]`, `List[List[str]]`, *optional*):
            The sequence or batch of sequences of entities to be encoded. Each sequence consists of strings
            representing entities, i.e., special entities (e.g., [MASK]) or entity titles of Wikipedia (e.g., Los
            Angeles). This argument is ignored if you specify the `task` argument in the forwardor. The length of
            each sequence must be equal to the length of each sequence of `entity_spans_pair`. If you specify
            `entity_spans_pair` without specifying this argument, the entity sequence or the batch of entity
            sequences is automatically forwarded by filling it with the [MASK] entity.
        max_entity_length (`int`, *optional*):
            The maximum length of `entity_ids`.
    """
    # Input type checking for clearer error
    is_valid_single_text = isinstance(text, str)
    is_valid_batch_text = isinstance(text, (list, tuple)) and (len(text) == 0 or (isinstance(text[0], str)))
    if not (is_valid_single_text or is_valid_batch_text):
        raise ValueError("text input must be of type `str` (single example) or `List[str]` (batch).")

    is_valid_single_text_pair = isinstance(text_pair, str)
    is_valid_batch_text_pair = isinstance(text_pair, (list, tuple)) and (
        len(text_pair) == 0 or isinstance(text_pair[0], str)
    )
    if not (text_pair is None or is_valid_single_text_pair or is_valid_batch_text_pair):
        raise ValueError("text_pair input must be of type `str` (single example) or `List[str]` (batch).")

    is_batched = bool(isinstance(text, (list, tuple)))

    if is_batched:
        batch_text_or_text_pairs = list(zip(text, text_pair)) if text_pair is not None else text
        if entities is None:
            batch_entities_or_entities_pairs = None
        else:
            batch_entities_or_entities_pairs = (
                list(zip(entities, entities_pair)) if entities_pair is not None else entities
            )

        if entity_spans is None:
            batch_entity_spans_or_entity_spans_pairs = None
        else:
            batch_entity_spans_or_entity_spans_pairs = (
                list(zip(entity_spans, entity_spans_pair)) if entity_spans_pair is not None else entity_spans
            )

        return self.batch_encode_plus(
            batch_text_or_text_pairs=batch_text_or_text_pairs,
            batch_entity_spans_or_entity_spans_pairs=batch_entity_spans_or_entity_spans_pairs,
            batch_entities_or_entities_pairs=batch_entities_or_entities_pairs,
            add_special_tokens=add_special_tokens,
            padding=padding,
            truncation=truncation,
            max_length=max_length,
            max_entity_length=max_entity_length,
            stride=stride,
            is_split_into_words=is_split_into_words,
            pad_to_multiple_of=pad_to_multiple_of,
            return_tensors=return_tensors,
            return_token_type_ids=return_token_type_ids,
            return_attention_mask=return_attention_mask,
            return_overflowing_tokens=return_overflowing_tokens,
            return_special_tokens_mask=return_special_tokens_mask,
            return_offsets_mapping=return_offsets_mapping,
            return_length=return_length,
            verbose=verbose,
            **kwargs,
        )
    return self.encode_plus(
        text=text,
        text_pair=text_pair,
        entity_spans=entity_spans,
        entity_spans_pair=entity_spans_pair,
        entities=entities,
        entities_pair=entities_pair,
        add_special_tokens=add_special_tokens,
        padding=padding,
        truncation=truncation,
        max_length=max_length,
        max_entity_length=max_entity_length,
        stride=stride,
        is_split_into_words=is_split_into_words,
        pad_to_multiple_of=pad_to_multiple_of,
        return_tensors=return_tensors,
        return_token_type_ids=return_token_type_ids,
        return_attention_mask=return_attention_mask,
        return_overflowing_tokens=return_overflowing_tokens,
        return_special_tokens_mask=return_special_tokens_mask,
        return_offsets_mapping=return_offsets_mapping,
        return_length=return_length,
        verbose=verbose,
        **kwargs,
    )

mindnlp.transformers.models.luke.tokenization_luke.LukeTokenizer.__init__(vocab_file, merges_file, entity_vocab_file, task=None, max_entity_length=32, max_mention_length=30, entity_token_1='<ent>', entity_token_2='<ent2>', entity_unk_token='[UNK]', entity_pad_token='[PAD]', entity_mask_token='[MASK]', entity_mask2_token='[MASK2]', errors='replace', bos_token='<s>', eos_token='</s>', sep_token='</s>', cls_token='<s>', unk_token='<unk>', pad_token='<pad>', mask_token='<mask>', add_prefix_space=False, **kwargs)

Initialize the LukeTokenizer class.

This method initializes an instance of the LukeTokenizer class. It takes the following parameters:

PARAMETER DESCRIPTION
self

The instance of the class.
vocab_file

The path to the vocabulary file.

TYPE: str

merges_file

The path to the merges file.

TYPE: str

entity_vocab_file

The path to the entity vocabulary file.

TYPE: str

task

The task for which the tokenizer is used. Defaults to None.

TYPE: str DEFAULT: None

max_entity_length

The maximum length of the entity. Defaults to 32.

TYPE: int DEFAULT: 32

max_mention_length

The maximum length of the mention. Defaults to 30.

TYPE: int DEFAULT: 30

entity_token_1

The first entity token. Defaults to ''.

TYPE: str DEFAULT: '<ent>'

entity_token_2

The second entity token. Defaults to ''.

TYPE: str DEFAULT: '<ent2>'

entity_unk_token

The unknown entity token. Defaults to '[UNK]'.

TYPE: str DEFAULT: '[UNK]'

entity_pad_token

The padding entity token. Defaults to '[PAD]'.

TYPE: str DEFAULT: '[PAD]'

entity_mask_token

The masked entity token. Defaults to '[MASK]'.

TYPE: str DEFAULT: '[MASK]'

entity_mask2_token

The second masked entity token. Defaults to '[MASK2]'.

TYPE: str DEFAULT: '[MASK2]'

errors

The error handling strategy. Defaults to 'replace'.

TYPE: str DEFAULT: 'replace'

bos_token

The beginning of sentence token. Defaults to ''.

TYPE: str DEFAULT: '<s>'

eos_token

The end of sentence token. Defaults to ''.

TYPE: str DEFAULT: '</s>'

sep_token

The separator token. Defaults to ''.

TYPE: str DEFAULT: '</s>'

cls_token

The classification token. Defaults to ''.

TYPE: str DEFAULT: '<s>'

unk_token

The unknown token. Defaults to ''.

TYPE: str DEFAULT: '<unk>'

pad_token

The padding token. Defaults to ''.

TYPE: str DEFAULT: '<pad>'

mask_token

The masked token. Defaults to ''.

TYPE: str DEFAULT: '<mask>'

add_prefix_space

Whether to add space before the token. Defaults to False.

TYPE: bool DEFAULT: False

RETURNS DESCRIPTION

None
RAISES DESCRIPTION
ValueError

If the specified entity special token is not found in the entity vocabulary file.
ValueError

If the task is not supported. Select task from ['entity_classification', 'entity_pair_classification', 'entity_span_classification'] only.
Source code in mindnlp\transformers\models\luke\tokenization_luke.py 
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
def __init__(
    self,
    vocab_file,
    merges_file,
    entity_vocab_file,
    task=None,
    max_entity_length=32,
    max_mention_length=30,
    entity_token_1="<ent>",
    entity_token_2="<ent2>",
    entity_unk_token="[UNK]",
    entity_pad_token="[PAD]",
    entity_mask_token="[MASK]",
    entity_mask2_token="[MASK2]",
    errors="replace",
    bos_token="<s>",
    eos_token="</s>",
    sep_token="</s>",
    cls_token="<s>",
    unk_token="<unk>",
    pad_token="<pad>",
    mask_token="<mask>",
    add_prefix_space=False,
    **kwargs,
):
    """Initialize the LukeTokenizer class.

    This method initializes an instance of the LukeTokenizer class. It takes the following parameters:

    Args:
        self: The instance of the class.
        vocab_file (str): The path to the vocabulary file.
        merges_file (str): The path to the merges file.
        entity_vocab_file (str): The path to the entity vocabulary file.
        task (str, optional): The task for which the tokenizer is used. Defaults to None.
        max_entity_length (int, optional): The maximum length of the entity. Defaults to 32.
        max_mention_length (int, optional): The maximum length of the mention. Defaults to 30.
        entity_token_1 (str, optional): The first entity token. Defaults to '<ent>'.
        entity_token_2 (str, optional): The second entity token. Defaults to '<ent2>'.
        entity_unk_token (str, optional): The unknown entity token. Defaults to '[UNK]'.
        entity_pad_token (str, optional): The padding entity token. Defaults to '[PAD]'.
        entity_mask_token (str, optional): The masked entity token. Defaults to '[MASK]'.
        entity_mask2_token (str, optional): The second masked entity token. Defaults to '[MASK2]'.
        errors (str, optional): The error handling strategy. Defaults to 'replace'.
        bos_token (str, optional): The beginning of sentence token. Defaults to '<s>'.
        eos_token (str, optional): The end of sentence token. Defaults to '</s>'.
        sep_token (str, optional): The separator token. Defaults to '</s>'.
        cls_token (str, optional): The classification token. Defaults to '<s>'.
        unk_token (str, optional): The unknown token. Defaults to '<unk>'.
        pad_token (str, optional): The padding token. Defaults to '<pad>'.
        mask_token (str, optional): The masked token. Defaults to '<mask>'.
        add_prefix_space (bool, optional): Whether to add space before the token. Defaults to False.

    Returns:
        None

    Raises:
        ValueError: If the specified entity special token is not found in the entity vocabulary file.
        ValueError: If the task is not supported. Select task from ['entity_classification',
            'entity_pair_classification', 'entity_span_classification'] only.
    """
    bos_token = AddedToken(bos_token, lstrip=False, rstrip=False) if isinstance(bos_token, str) else bos_token
    eos_token = AddedToken(eos_token, lstrip=False, rstrip=False) if isinstance(eos_token, str) else eos_token
    sep_token = AddedToken(sep_token, lstrip=False, rstrip=False) if isinstance(sep_token, str) else sep_token
    cls_token = AddedToken(cls_token, lstrip=False, rstrip=False) if isinstance(cls_token, str) else cls_token
    unk_token = AddedToken(unk_token, lstrip=False, rstrip=False) if isinstance(unk_token, str) else unk_token
    pad_token = AddedToken(pad_token, lstrip=False, rstrip=False) if isinstance(pad_token, str) else pad_token

    # Mask token behave like a normal word, i.e. include the space before it
    mask_token = AddedToken(mask_token, lstrip=True, rstrip=False) if isinstance(mask_token, str) else mask_token

    with open(vocab_file, encoding="utf-8") as vocab_handle:
        self.encoder = json.load(vocab_handle)
    self.decoder = {v: k for k, v in self.encoder.items()}
    self.errors = errors  # how to handle errors in decoding
    self.byte_encoder = bytes_to_unicode()
    self.byte_decoder = {v: k for k, v in self.byte_encoder.items()}
    with open(merges_file, encoding="utf-8") as merges_handle:
        bpe_merges = merges_handle.read().split("\n")[1:-1]
    bpe_merges = [tuple(merge.split()) for merge in bpe_merges]
    self.bpe_ranks = dict(zip(bpe_merges, range(len(bpe_merges))))
    self.cache = {}
    self.add_prefix_space = add_prefix_space

    # Should have added re.IGNORECASE so BPE merges can happen for capitalized versions of contractions
    self.pat = re.compile(r"""'s|'t|'re|'ve|'m|'ll|'d| ?\p{L}+| ?\p{N}+| ?[^\s\p{L}\p{N}]+|\s+(?!\S)|\s+""")

    # we add 2 special tokens for downstream tasks
    # for more information about lstrip and rstrip, see https://github.com/huggingface/transformers/pull/2778
    entity_token_1 = (
        AddedToken(entity_token_1, lstrip=False, rstrip=False)
        if isinstance(entity_token_1, str)
        else entity_token_1
    )
    entity_token_2 = (
        AddedToken(entity_token_2, lstrip=False, rstrip=False)
        if isinstance(entity_token_2, str)
        else entity_token_2
    )
    kwargs["additional_special_tokens"] = kwargs.get("additional_special_tokens", [])
    kwargs["additional_special_tokens"] += [entity_token_1, entity_token_2]

    with open(entity_vocab_file, encoding="utf-8") as entity_vocab_handle:
        self.entity_vocab = json.load(entity_vocab_handle)
    for entity_special_token in [entity_unk_token, entity_pad_token, entity_mask_token, entity_mask2_token]:
        if entity_special_token not in self.entity_vocab:
            raise ValueError(
                f"Specified entity special token ``{entity_special_token}`` is not found in entity_vocab. "
                f"Probably an incorrect entity vocab file is loaded: {entity_vocab_file}."
            )
    self.entity_unk_token_id = self.entity_vocab[entity_unk_token]
    self.entity_pad_token_id = self.entity_vocab[entity_pad_token]
    self.entity_mask_token_id = self.entity_vocab[entity_mask_token]
    self.entity_mask2_token_id = self.entity_vocab[entity_mask2_token]

    self.task = task
    if task is None or task == "entity_span_classification":
        self.max_entity_length = max_entity_length
    elif task == "entity_classification":
        self.max_entity_length = 1
    elif task == "entity_pair_classification":
        self.max_entity_length = 2
    else:
        raise ValueError(
            f"Task {task} not supported. Select task from ['entity_classification', 'entity_pair_classification',"
            " 'entity_span_classification'] only."
        )

    self.max_mention_length = max_mention_length

    super().__init__(
        errors=errors,
        bos_token=bos_token,
        eos_token=eos_token,
        unk_token=unk_token,
        sep_token=sep_token,
        cls_token=cls_token,
        pad_token=pad_token,
        mask_token=mask_token,
        add_prefix_space=add_prefix_space,
        task=task,
        max_entity_length=32,
        max_mention_length=30,
        entity_token_1="<ent>",
        entity_token_2="<ent2>",
        entity_unk_token=entity_unk_token,
        entity_pad_token=entity_pad_token,
        entity_mask_token=entity_mask_token,
        entity_mask2_token=entity_mask2_token,
        **kwargs,
    )

mindnlp.transformers.models.luke.tokenization_luke.LukeTokenizer.bpe(token)

This method 'bpe' in the class 'LukeTokenizer' performs Byte Pair Encoding (BPE) on the input token.

PARAMETER DESCRIPTION
self

The instance of the LukeTokenizer class.

TYPE: LukeTokenizer

token

The input token to be processed using Byte Pair Encoding.

TYPE: str

RETURNS DESCRIPTION
str

The processed token after applying Byte Pair Encoding.
RAISES DESCRIPTION
ValueError

If the input token is empty.
TypeError

If the input token is not a string.
Source code in mindnlp\transformers\models\luke\tokenization_luke.py 
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
def bpe(self, token):
    """
    This method 'bpe' in the class 'LukeTokenizer' performs Byte Pair Encoding (BPE) on the input token.

    Args:
        self (LukeTokenizer): The instance of the LukeTokenizer class.
        token (str): The input token to be processed using Byte Pair Encoding.

    Returns:
        str: The processed token after applying Byte Pair Encoding.

    Raises:
        ValueError: If the input token is empty.
        TypeError: If the input token is not a string.
    """
    if token in self.cache:
        return self.cache[token]
    word = tuple(token)
    pairs = get_pairs(word)

    if not pairs:
        return token

    while True:
        bigram = min(pairs, key=lambda pair: self.bpe_ranks.get(pair, float("inf")))
        if bigram not in self.bpe_ranks:
            break
        first, second = bigram
        new_word = []
        i = 0
        while i < len(word):
            try:
                j = word.index(first, i)
            except ValueError:
                new_word.extend(word[i:])
                break
            else:
                new_word.extend(word[i:j])
                i = j

            if word[i] == first and i < len(word) - 1 and word[i + 1] == second:
                new_word.append(first + second)
                i += 2
            else:
                new_word.append(word[i])
                i += 1
        new_word = tuple(new_word)
        word = new_word
        if len(word) == 1:
            break
        pairs = get_pairs(word)
    word = " ".join(word)
    self.cache[token] = word
    return word

mindnlp.transformers.models.luke.tokenization_luke.LukeTokenizer.build_inputs_with_special_tokens(token_ids_0, token_ids_1=None)

Build model inputs from a sequence or a pair of sequence for sequence classification tasks by concatenating and adding special tokens. A LUKE sequence has the following format:

  • single sequence: <s> X </s>
  • pair of sequences: <s> A </s></s> B </s>
PARAMETER DESCRIPTION
token_ids_0

List of IDs to which the special tokens will be added.

TYPE: `List[int]`

token_ids_1

Optional second list of IDs for sequence pairs.

TYPE: `List[int]`, *optional* DEFAULT: None

RETURNS DESCRIPTION
List[int]

List[int]: List of input IDs with the appropriate special tokens.
Source code in mindnlp\transformers\models\luke\tokenization_luke.py 
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
def build_inputs_with_special_tokens(
    self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None
) -> List[int]:
    """
    Build model inputs from a sequence or a pair of sequence for sequence classification tasks by concatenating and
    adding special tokens. A LUKE sequence has the following format:

    - single sequence: `<s> X </s>`
    - pair of sequences: `<s> A </s></s> B </s>`

    Args:
        token_ids_0 (`List[int]`):
            List of IDs to which the special tokens will be added.
        token_ids_1 (`List[int]`, *optional*):
            Optional second list of IDs for sequence pairs.

    Returns:
        `List[int]`: List of [input IDs](../glossary#input-ids) with the appropriate special tokens.
    """
    if token_ids_1 is None:
        return [self.cls_token_id] + token_ids_0 + [self.sep_token_id]
    cls = [self.cls_token_id]
    sep = [self.sep_token_id]
    return cls + token_ids_0 + sep + sep + token_ids_1 + sep

mindnlp.transformers.models.luke.tokenization_luke.LukeTokenizer.convert_tokens_to_string(tokens)

Converts a sequence of tokens (string) in a single string.

Source code in mindnlp\transformers\models\luke\tokenization_luke.py 
558
559
560
561
562
def convert_tokens_to_string(self, tokens):
    """Converts a sequence of tokens (string) in a single string."""
    text = "".join(tokens)
    text = bytearray([self.byte_decoder[c] for c in text]).decode("utf-8", errors=self.errors)
    return text

mindnlp.transformers.models.luke.tokenization_luke.LukeTokenizer.create_token_type_ids_from_sequences(token_ids_0, token_ids_1=None)

Create a mask from the two sequences passed to be used in a sequence-pair classification task. LUKE does not make use of token type ids, therefore a list of zeros is returned.

PARAMETER DESCRIPTION
token_ids_0

List of IDs.

TYPE: `List[int]`

token_ids_1

Optional second list of IDs for sequence pairs.

TYPE: `List[int]`, *optional* DEFAULT: None

RETURNS DESCRIPTION
List[int]

List[int]: List of zeros.
Source code in mindnlp\transformers\models\luke\tokenization_luke.py 
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
def create_token_type_ids_from_sequences(
    self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None
) -> List[int]:
    """
    Create a mask from the two sequences passed to be used in a sequence-pair classification task. LUKE does not
    make use of token type ids, therefore a list of zeros is returned.

    Args:
        token_ids_0 (`List[int]`):
            List of IDs.
        token_ids_1 (`List[int]`, *optional*):
            Optional second list of IDs for sequence pairs.

    Returns:
        `List[int]`: List of zeros.
    """
    sep = [self.sep_token_id]
    cls = [self.cls_token_id]

    if token_ids_1 is None:
        return len(cls + token_ids_0 + sep) * [0]
    return len(cls + token_ids_0 + sep + sep + token_ids_1 + sep) * [0]

mindnlp.transformers.models.luke.tokenization_luke.LukeTokenizer.get_special_tokens_mask(token_ids_0, token_ids_1=None, already_has_special_tokens=False)

Retrieve sequence ids from a token list that has no special tokens added. This method is called when adding special tokens using the tokenizer prepare_for_model method.

PARAMETER DESCRIPTION
token_ids_0

List of IDs.

TYPE: `List[int]`

token_ids_1

Optional second list of IDs for sequence pairs.

TYPE: `List[int]`, *optional* DEFAULT: None

already_has_special_tokens

Whether or not the token list is already formatted with special tokens for the model.

TYPE: `bool`, *optional*, defaults to `False` DEFAULT: False

RETURNS DESCRIPTION
List[int]

List[int]: A list of integers in the range [0, 1]: 1 for a special token, 0 for a sequence token.
Source code in mindnlp\transformers\models\luke\tokenization_luke.py 
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
def get_special_tokens_mask(
    self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None, already_has_special_tokens: bool = False
) -> List[int]:
    """
    Retrieve sequence ids from a token list that has no special tokens added. This method is called when adding
    special tokens using the tokenizer `prepare_for_model` method.

    Args:
        token_ids_0 (`List[int]`):
            List of IDs.
        token_ids_1 (`List[int]`, *optional*):
            Optional second list of IDs for sequence pairs.
        already_has_special_tokens (`bool`, *optional*, defaults to `False`):
            Whether or not the token list is already formatted with special tokens for the model.

    Returns:
        `List[int]`: A list of integers in the range [0, 1]: 1 for a special token, 0 for a sequence token.
    """
    if already_has_special_tokens:
        return super().get_special_tokens_mask(
            token_ids_0=token_ids_0, token_ids_1=token_ids_1, already_has_special_tokens=True
        )

    if token_ids_1 is None:
        return [1] + ([0] * len(token_ids_0)) + [1]
    return [1] + ([0] * len(token_ids_0)) + [1, 1] + ([0] * len(token_ids_1)) + [1]

mindnlp.transformers.models.luke.tokenization_luke.LukeTokenizer.get_vocab()

Retrieves the vocabulary dictionary for the 'LukeTokenizer' class.

PARAMETER DESCRIPTION
self

An instance of the 'LukeTokenizer' class.
RETURNS DESCRIPTION
dict

A dictionary containing the vocabulary of the tokenizer. The keys are the tokens and the values are their corresponding IDs.
Source code in mindnlp\transformers\models\luke\tokenization_luke.py 
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
def get_vocab(self):
    """
    Retrieves the vocabulary dictionary for the 'LukeTokenizer' class.

    Args:
        self: An instance of the 'LukeTokenizer' class.

    Returns:
        dict: A dictionary containing the vocabulary of the tokenizer. The keys are the tokens
            and the values are their corresponding IDs.

    Raises:
        None.
    """
    vocab = dict(self.encoder).copy()
    vocab.update(self.added_tokens_encoder)
    return vocab

mindnlp.transformers.models.luke.tokenization_luke.LukeTokenizer.pad(encoded_inputs, padding=True, max_length=None, max_entity_length=None, pad_to_multiple_of=None, return_attention_mask=None, return_tensors=None, verbose=True)

Pad a single encoded input or a batch of encoded inputs up to predefined length or to the max sequence length in the batch. Padding side (left/right) padding token ids are defined at the tokenizer level (with self.padding_side, self.pad_token_id and self.pad_token_type_id) .. note:: If the encoded_inputs passed are dictionary of numpy arrays, PyTorch tensors or TensorFlow tensors, the result will use the same type unless you provide a different tensor type with return_tensors. In the case of PyTorch tensors, you will lose the specific device of your tensors however.

PARAMETER DESCRIPTION
encoded_inputs

Tokenized inputs. Can represent one input ([BatchEncoding] or Dict[str, List[int]]) or a batch of tokenized inputs (list of [BatchEncoding], Dict[str, List[List[int]]] or List[Dict[str, List[int]]]) so you can use this method during preprocessing as well as in a PyTorch Dataloader collate function. Instead of List[int] you can have tensors (numpy arrays, PyTorch tensors or TensorFlow tensors), see the note above for the return type.

TYPE: [`BatchEncoding`], list of [`BatchEncoding`], `Dict[str, List[int]]`, `Dict[str, List[List[int]]` or `List[Dict[str, List[int]]]`

padding

Select a strategy to pad the returned sequences (according to the model's padding side and padding index) among:

  • True or 'longest': Pad to the longest sequence in the batch (or no padding if only a single sequence if provided).
  • 'max_length': Pad to a maximum length specified with the argument max_length or to the maximum acceptable input length for the model if that argument is not provided.
  • False or 'do_not_pad' (default): No padding (i.e., can output a batch with sequences of different lengths).

TYPE: `bool`, `str` or [`~utils.PaddingStrategy`], *optional*, defaults to `True` DEFAULT: True

max_length

Maximum length of the returned list and optionally padding length (see above).

TYPE: `int`, *optional* DEFAULT: None

max_entity_length

The maximum length of the entity sequence.

TYPE: `int`, *optional* DEFAULT: None

pad_to_multiple_of

If set will pad the sequence to a multiple of the provided value. This is especially useful to enable the use of Tensor Cores on NVIDIA hardware with compute capability >= 7.5 (Volta).

TYPE: `int`, *optional* DEFAULT: None

return_attention_mask

Whether to return the attention mask. If left to the default, will return the attention mask according to the specific tokenizer's default, defined by the return_outputs attribute. What are attention masks?

TYPE: `bool`, *optional* DEFAULT: None

return_tensors

If set, will return tensors instead of list of python integers. Acceptable values are:

  • 'tf': Return TensorFlow tf.constant objects.
  • 'pt': Return PyTorch torch.Tensor objects.
  • 'np': Return Numpy np.ndarray objects.

TYPE: `str` or [`~utils.TensorType`], *optional* DEFAULT: None

verbose

Whether or not to print more information and warnings.

TYPE: `bool`, *optional*, defaults to `True` DEFAULT: True

Source code in mindnlp\transformers\models\luke\tokenization_luke.py 
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
def pad(
    self,
    encoded_inputs: Union[
        BatchEncoding,
        List[BatchEncoding],
        Dict[str, EncodedInput],
        Dict[str, List[EncodedInput]],
        List[Dict[str, EncodedInput]],
    ],
    padding: Union[bool, str, PaddingStrategy] = True,
    max_length: Optional[int] = None,
    max_entity_length: Optional[int] = None,
    pad_to_multiple_of: Optional[int] = None,
    return_attention_mask: Optional[bool] = None,
    return_tensors: Optional[Union[str, TensorType]] = None,
    verbose: bool = True,
) -> BatchEncoding:
    """
    Pad a single encoded input or a batch of encoded inputs up to predefined length or to the max sequence length
    in the batch. Padding side (left/right) padding token ids are defined at the tokenizer level (with
    `self.padding_side`, `self.pad_token_id` and `self.pad_token_type_id`) .. note:: If the `encoded_inputs` passed
    are dictionary of numpy arrays, PyTorch tensors or TensorFlow tensors, the result will use the same type unless
    you provide a different tensor type with `return_tensors`. In the case of PyTorch tensors, you will lose the
    specific device of your tensors however.

    Args:
        encoded_inputs ([`BatchEncoding`], list of [`BatchEncoding`], `Dict[str, List[int]]`, `Dict[str, List[List[int]]` or `List[Dict[str, List[int]]]`):
            Tokenized inputs. Can represent one input ([`BatchEncoding`] or `Dict[str, List[int]]`) or a batch of
            tokenized inputs (list of [`BatchEncoding`], *Dict[str, List[List[int]]]* or *List[Dict[str,
            List[int]]]*) so you can use this method during preprocessing as well as in a PyTorch Dataloader
            collate function. Instead of `List[int]` you can have tensors (numpy arrays, PyTorch tensors or
            TensorFlow tensors), see the note above for the return type.
        padding (`bool`, `str` or [`~utils.PaddingStrategy`], *optional*, defaults to `True`):
             Select a strategy to pad the returned sequences (according to the model's padding side and padding
             index) among:

            - `True` or `'longest'`: Pad to the longest sequence in the batch (or no padding if only a single
            sequence if provided).
            - `'max_length'`: Pad to a maximum length specified with the argument `max_length` or to the maximum
            acceptable input length for the model if that argument is not provided.
            - `False` or `'do_not_pad'` (default): No padding (i.e., can output a batch with sequences of different
            lengths).
        max_length (`int`, *optional*):
            Maximum length of the returned list and optionally padding length (see above).
        max_entity_length (`int`, *optional*):
            The maximum length of the entity sequence.
        pad_to_multiple_of (`int`, *optional*):
            If set will pad the sequence to a multiple of the provided value. This is especially useful to enable
            the use of Tensor Cores on NVIDIA hardware with compute capability `>= 7.5` (Volta).
        return_attention_mask (`bool`, *optional*):
            Whether to return the attention mask. If left to the default, will return the attention mask according
            to the specific tokenizer's default, defined by the `return_outputs` attribute. [What are attention
            masks?](../glossary#attention-mask)
        return_tensors (`str` or [`~utils.TensorType`], *optional*):
            If set, will return tensors instead of list of python integers. Acceptable values are:

            - `'tf'`: Return TensorFlow `tf.constant` objects.
            - `'pt'`: Return PyTorch `torch.Tensor` objects.
            - `'np'`: Return Numpy `np.ndarray` objects.
        verbose (`bool`, *optional*, defaults to `True`):
            Whether or not to print more information and warnings.
    """
    # If we have a list of dicts, let's convert it in a dict of lists
    # We do this to allow using this method as a collate_fn function in PyTorch Dataloader
    if isinstance(encoded_inputs, (list, tuple)) and isinstance(encoded_inputs[0], Mapping):
        encoded_inputs = {key: [example[key] for example in encoded_inputs] for key in encoded_inputs[0].keys()}

    # The model's main input name, usually `input_ids`, has be passed for padding
    if self.model_input_names[0] not in encoded_inputs:
        raise ValueError(
            "You should supply an encoding or a list of encodings to this method "
            f"that includes {self.model_input_names[0]}, but you provided {list(encoded_inputs.keys())}"
        )

    required_input = encoded_inputs[self.model_input_names[0]]

    if not required_input:
        if return_attention_mask:
            encoded_inputs["attention_mask"] = []
        return encoded_inputs

    # If we have PyTorch/TF/NumPy tensors/arrays as inputs, we cast them as python objects
    # and rebuild them afterwards if no return_tensors is specified
    # Note that we lose the specific device the tensor may be on for PyTorch

    first_element = required_input[0]
    if isinstance(first_element, (list, tuple)):
        # first_element might be an empty list/tuple in some edge cases so we grab the first non empty element.
        index = 0
        while len(required_input[index]) == 0:
            index += 1
        if index < len(required_input):
            first_element = required_input[index][0]
    # At this state, if `first_element` is still a list/tuple, it's an empty one so there is nothing to do.
    if not isinstance(first_element, (int, list, tuple)):
        if is_mindspore_tensor(first_element):
            return_tensors = "ms" if return_tensors is None else return_tensors
        elif isinstance(first_element, np.ndarray):
            return_tensors = "np" if return_tensors is None else return_tensors
        else:
            raise ValueError(
                f"type of {first_element} unknown: {type(first_element)}. "
                "Should be one of a python, numpy, pytorch or tensorflow object."
            )

        for key, value in encoded_inputs.items():
            encoded_inputs[key] = to_py_obj(value)

    # Convert padding_strategy in PaddingStrategy
    padding_strategy, _, max_length, _ = self._get_padding_truncation_strategies(
        padding=padding, max_length=max_length, verbose=verbose
    )

    if max_entity_length is None:
        max_entity_length = self.max_entity_length

    required_input = encoded_inputs[self.model_input_names[0]]
    if required_input and not isinstance(required_input[0], (list, tuple)):
        encoded_inputs = self._pad(
            encoded_inputs,
            max_length=max_length,
            max_entity_length=max_entity_length,
            padding_strategy=padding_strategy,
            pad_to_multiple_of=pad_to_multiple_of,
            return_attention_mask=return_attention_mask,
        )
        return BatchEncoding(encoded_inputs, tensor_type=return_tensors)

    batch_size = len(required_input)
    if any(len(v) != batch_size for v in encoded_inputs.values()):
        raise ValueError("Some items in the output dictionary have a different batch size than others.")

    if padding_strategy == PaddingStrategy.LONGEST:
        max_length = max(len(inputs) for inputs in required_input)
        max_entity_length = (
            max(len(inputs) for inputs in encoded_inputs["entity_ids"]) if "entity_ids" in encoded_inputs else 0
        )
        padding_strategy = PaddingStrategy.MAX_LENGTH

    batch_outputs = {}
    for i in range(batch_size):
        inputs = {k: v[i] for k, v in encoded_inputs.items()}
        outputs = self._pad(
            inputs,
            max_length=max_length,
            max_entity_length=max_entity_length,
            padding_strategy=padding_strategy,
            pad_to_multiple_of=pad_to_multiple_of,
            return_attention_mask=return_attention_mask,
        )

        for key, value in outputs.items():
            if key not in batch_outputs:
                batch_outputs[key] = []
            batch_outputs[key].append(value)

    return BatchEncoding(batch_outputs, tensor_type=return_tensors)

mindnlp.transformers.models.luke.tokenization_luke.LukeTokenizer.prepare_for_model(ids, pair_ids=None, entity_ids=None, pair_entity_ids=None, entity_token_spans=None, pair_entity_token_spans=None, add_special_tokens=True, padding=False, truncation=None, max_length=None, max_entity_length=None, stride=0, pad_to_multiple_of=None, return_tensors=None, return_token_type_ids=None, return_attention_mask=None, return_overflowing_tokens=False, return_special_tokens_mask=False, return_offsets_mapping=False, return_length=False, verbose=True, prepend_batch_axis=False, **kwargs)

Prepares a sequence of input id, entity id and entity span, or a pair of sequences of inputs ids, entity ids, entity spans so that it can be used by the model. It adds special tokens, truncates sequences if overflowing while taking into account the special tokens and manages a moving window (with user defined stride) for overflowing tokens. Please Note, for pair_ids different than None and truncation_strategy = longest_first or True, it is not possible to return overflowing tokens. Such a combination of arguments will raise an error.

PARAMETER DESCRIPTION
ids

Tokenized input ids of the first sequence.

TYPE: `List[int]`

pair_ids

Tokenized input ids of the second sequence.

TYPE: `List[int]`, *optional* DEFAULT: None

entity_ids

Entity ids of the first sequence.

TYPE: `List[int]`, *optional* DEFAULT: None

pair_entity_ids

Entity ids of the second sequence.

TYPE: `List[int]`, *optional* DEFAULT: None

entity_token_spans

Entity spans of the first sequence.

TYPE: `List[Tuple[int, int]]`, *optional* DEFAULT: None

pair_entity_token_spans

Entity spans of the second sequence.

TYPE: `List[Tuple[int, int]]`, *optional* DEFAULT: None

max_entity_length

The maximum length of the entity sequence.

TYPE: `int`, *optional* DEFAULT: None

Source code in mindnlp\transformers\models\luke\tokenization_luke.py 
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
def prepare_for_model(
    self,
    ids: List[int],
    pair_ids: Optional[List[int]] = None,
    entity_ids: Optional[List[int]] = None,
    pair_entity_ids: Optional[List[int]] = None,
    entity_token_spans: Optional[List[Tuple[int, int]]] = None,
    pair_entity_token_spans: Optional[List[Tuple[int, int]]] = None,
    add_special_tokens: bool = True,
    padding: Union[bool, str, PaddingStrategy] = False,
    truncation: Union[bool, str, TruncationStrategy] = None,
    max_length: Optional[int] = None,
    max_entity_length: Optional[int] = None,
    stride: int = 0,
    pad_to_multiple_of: Optional[int] = None,
    return_tensors: Optional[Union[str, TensorType]] = None,
    return_token_type_ids: Optional[bool] = None,
    return_attention_mask: Optional[bool] = None,
    return_overflowing_tokens: bool = False,
    return_special_tokens_mask: bool = False,
    return_offsets_mapping: bool = False,
    return_length: bool = False,
    verbose: bool = True,
    prepend_batch_axis: bool = False,
    **kwargs,
) -> BatchEncoding:
    """
    Prepares a sequence of input id, entity id and entity span, or a pair of sequences of inputs ids, entity ids,
    entity spans so that it can be used by the model. It adds special tokens, truncates sequences if overflowing
    while taking into account the special tokens and manages a moving window (with user defined stride) for
    overflowing tokens. Please Note, for *pair_ids* different than `None` and *truncation_strategy = longest_first*
    or `True`, it is not possible to return overflowing tokens. Such a combination of arguments will raise an
    error.

    Args:
        ids (`List[int]`):
            Tokenized input ids of the first sequence.
        pair_ids (`List[int]`, *optional*):
            Tokenized input ids of the second sequence.
        entity_ids (`List[int]`, *optional*):
            Entity ids of the first sequence.
        pair_entity_ids (`List[int]`, *optional*):
            Entity ids of the second sequence.
        entity_token_spans (`List[Tuple[int, int]]`, *optional*):
            Entity spans of the first sequence.
        pair_entity_token_spans (`List[Tuple[int, int]]`, *optional*):
            Entity spans of the second sequence.
        max_entity_length (`int`, *optional*):
            The maximum length of the entity sequence.
    """
    # Backward compatibility for 'truncation_strategy', 'pad_to_max_length'
    padding_strategy, truncation_strategy, max_length, kwargs = self._get_padding_truncation_strategies(
        padding=padding,
        truncation=truncation,
        max_length=max_length,
        pad_to_multiple_of=pad_to_multiple_of,
        verbose=verbose,
        **kwargs,
    )

    # Compute lengths
    pair = bool(pair_ids is not None)
    len_ids = len(ids)
    len_pair_ids = len(pair_ids) if pair else 0

    if return_token_type_ids and not add_special_tokens:
        raise ValueError(
            "Asking to return token_type_ids while setting add_special_tokens to False "
            "results in an undefined behavior. Please set add_special_tokens to True or "
            "set return_token_type_ids to None."
        )
    if (
        return_overflowing_tokens
        and truncation_strategy == TruncationStrategy.LONGEST_FIRST
        and pair_ids is not None
    ):
        raise ValueError(
            "Not possible to return overflowing tokens for pair of sequences with the "
            "`longest_first`. Please select another truncation strategy than `longest_first`, "
            "for instance `only_second` or `only_first`."
        )

    # Load from model defaults
    if return_token_type_ids is None:
        return_token_type_ids = "token_type_ids" in self.model_input_names
    if return_attention_mask is None:
        return_attention_mask = "attention_mask" in self.model_input_names

    encoded_inputs = {}

    # Compute the total size of the returned word encodings
    total_len = len_ids + len_pair_ids + (self.num_special_tokens_to_add(pair=pair) if add_special_tokens else 0)

    # Truncation: Handle max sequence length and max_entity_length
    overflowing_tokens = []
    if truncation_strategy != TruncationStrategy.DO_NOT_TRUNCATE and max_length and total_len > max_length:
        # truncate words up to max_length
        ids, pair_ids, overflowing_tokens = self.truncate_sequences(
            ids,
            pair_ids=pair_ids,
            num_tokens_to_remove=total_len - max_length,
            truncation_strategy=truncation_strategy,
            stride=stride,
        )

    if return_overflowing_tokens:
        encoded_inputs["overflowing_tokens"] = overflowing_tokens
        encoded_inputs["num_truncated_tokens"] = total_len - max_length

    # Add special tokens
    if add_special_tokens:
        sequence = self.build_inputs_with_special_tokens(ids, pair_ids)
        token_type_ids = self.create_token_type_ids_from_sequences(ids, pair_ids)
        entity_token_offset = 1  # 1 * <s> token
        pair_entity_token_offset = len(ids) + 3  # 1 * <s> token & 2 * <sep> tokens
    else:
        sequence = ids + pair_ids if pair else ids
        token_type_ids = [0] * len(ids) + ([0] * len(pair_ids) if pair else [])
        entity_token_offset = 0
        pair_entity_token_offset = len(ids)

    # Build output dictionary
    encoded_inputs["input_ids"] = sequence
    if return_token_type_ids:
        encoded_inputs["token_type_ids"] = token_type_ids
    if return_special_tokens_mask:
        if add_special_tokens:
            encoded_inputs["special_tokens_mask"] = self.get_special_tokens_mask(ids, pair_ids)
        else:
            encoded_inputs["special_tokens_mask"] = [0] * len(sequence)

    # Set max entity length
    if not max_entity_length:
        max_entity_length = self.max_entity_length

    if entity_ids is not None:
        total_entity_len = 0
        num_invalid_entities = 0
        valid_entity_ids = [ent_id for ent_id, span in zip(entity_ids, entity_token_spans) if span[1] <= len(ids)]
        valid_entity_token_spans = [span for span in entity_token_spans if span[1] <= len(ids)]

        total_entity_len += len(valid_entity_ids)
        num_invalid_entities += len(entity_ids) - len(valid_entity_ids)

        valid_pair_entity_ids, valid_pair_entity_token_spans = None, None
        if pair_entity_ids is not None:
            valid_pair_entity_ids = [
                ent_id
                for ent_id, span in zip(pair_entity_ids, pair_entity_token_spans)
                if span[1] <= len(pair_ids)
            ]
            valid_pair_entity_token_spans = [span for span in pair_entity_token_spans if span[1] <= len(pair_ids)]
            total_entity_len += len(valid_pair_entity_ids)
            num_invalid_entities += len(pair_entity_ids) - len(valid_pair_entity_ids)

        if num_invalid_entities != 0:
            logger.warning(
                f"{num_invalid_entities} entities are ignored because their entity spans are invalid due to the"
                " truncation of input tokens"
            )

        if truncation_strategy != TruncationStrategy.DO_NOT_TRUNCATE and total_entity_len > max_entity_length:
            # truncate entities up to max_entity_length
            valid_entity_ids, valid_pair_entity_ids, overflowing_entities = self.truncate_sequences(
                valid_entity_ids,
                pair_ids=valid_pair_entity_ids,
                num_tokens_to_remove=total_entity_len - max_entity_length,
                truncation_strategy=truncation_strategy,
                stride=stride,
            )
            valid_entity_token_spans = valid_entity_token_spans[: len(valid_entity_ids)]
            if valid_pair_entity_token_spans is not None:
                valid_pair_entity_token_spans = valid_pair_entity_token_spans[: len(valid_pair_entity_ids)]

        if return_overflowing_tokens:
            encoded_inputs["overflowing_entities"] = overflowing_entities
            encoded_inputs["num_truncated_entities"] = total_entity_len - max_entity_length

        final_entity_ids = valid_entity_ids + valid_pair_entity_ids if valid_pair_entity_ids else valid_entity_ids
        encoded_inputs["entity_ids"] = list(final_entity_ids)
        entity_position_ids = []
        entity_start_positions = []
        entity_end_positions = []
        for token_spans, offset in (
            (valid_entity_token_spans, entity_token_offset),
            (valid_pair_entity_token_spans, pair_entity_token_offset),
        ):
            if token_spans is not None:
                for start, end in token_spans:
                    start += offset
                    end += offset
                    position_ids = list(range(start, end))[: self.max_mention_length]
                    position_ids += [-1] * (self.max_mention_length - end + start)
                    entity_position_ids.append(position_ids)
                    entity_start_positions.append(start)
                    entity_end_positions.append(end - 1)

        encoded_inputs["entity_position_ids"] = entity_position_ids
        if self.task == "entity_span_classification":
            encoded_inputs["entity_start_positions"] = entity_start_positions
            encoded_inputs["entity_end_positions"] = entity_end_positions

        if return_token_type_ids:
            encoded_inputs["entity_token_type_ids"] = [0] * len(encoded_inputs["entity_ids"])

    # Check lengths
    self._eventual_warn_about_too_long_sequence(encoded_inputs["input_ids"], max_length, verbose)

    # Padding
    if padding_strategy != PaddingStrategy.DO_NOT_PAD or return_attention_mask:
        encoded_inputs = self.pad(
            encoded_inputs,
            max_length=max_length,
            max_entity_length=max_entity_length,
            padding=padding_strategy.value,
            pad_to_multiple_of=pad_to_multiple_of,
            return_attention_mask=return_attention_mask,
        )

    if return_length:
        encoded_inputs["length"] = len(encoded_inputs["input_ids"])

    batch_outputs = BatchEncoding(
        encoded_inputs, tensor_type=return_tensors, prepend_batch_axis=prepend_batch_axis
    )

    return batch_outputs

mindnlp.transformers.models.luke.tokenization_luke.LukeTokenizer.prepare_for_tokenization(text, is_split_into_words=False, **kwargs)

Prepares the input text for tokenization by adding a prefix space if necessary.

PARAMETER DESCRIPTION
self

An instance of the LukeTokenizer class.

TYPE: LukeTokenizer

text

The input text to be tokenized.

TYPE: str

is_split_into_words

A flag indicating if the input text is already split into words. Defaults to False.

TYPE: bool DEFAULT: False

RETURNS DESCRIPTION
None

The method modifies the input text in-place.
Source code in mindnlp\transformers\models\luke\tokenization_luke.py 
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
def prepare_for_tokenization(self, text, is_split_into_words=False, **kwargs):
    """
    Prepares the input text for tokenization by adding a prefix space if necessary.

    Args:
        self (LukeTokenizer): An instance of the LukeTokenizer class.
        text (str): The input text to be tokenized.
        is_split_into_words (bool): A flag indicating if the input text is already split into words.
            Defaults to False.

    Returns:
        None: The method modifies the input text in-place.

    Raises:
        None.
    """
    add_prefix_space = kwargs.pop("add_prefix_space", self.add_prefix_space)
    if (is_split_into_words or add_prefix_space) and (len(text) > 0 and not text[0].isspace()):
        text = " " + text
    return (text, kwargs)

mindnlp.transformers.models.luke.tokenization_luke.LukeTokenizer.save_vocabulary(save_directory, filename_prefix=None)

Save the vocabulary to specified directory with an optional filename prefix.

PARAMETER DESCRIPTION
self

Instance of LukeTokenizer class.
save_directory

The directory path where the vocabulary files will be saved.

TYPE: str

filename_prefix

An optional prefix to be added to the filename. Default is None.

TYPE: Optional[str] DEFAULT: None

RETURNS DESCRIPTION
Tuple[str]

Tuple[str]: A tuple containing paths to the saved vocabulary files - vocab_file, merge_file, and entity_vocab_file.
RAISES DESCRIPTION
FileNotFoundError

If the specified save_directory does not exist.
IOError

If there is an issue with reading or writing the vocabulary files.
ValueError

If the provided filename_prefix is not a string.
Exception

Any other unexpected error that may occur during the execution of the method.
Source code in mindnlp\transformers\models\luke\tokenization_luke.py 
1898
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
def save_vocabulary(self, save_directory: str, filename_prefix: Optional[str] = None) -> Tuple[str]:
    """
    Save the vocabulary to specified directory with an optional filename prefix.

    Args:
        self: Instance of LukeTokenizer class.
        save_directory (str): The directory path where the vocabulary files will be saved.
        filename_prefix (Optional[str]): An optional prefix to be added to the filename. Default is None.

    Returns:
        Tuple[str]: A tuple containing paths to the saved vocabulary files - vocab_file, merge_file,
            and entity_vocab_file.

    Raises:
        FileNotFoundError: If the specified save_directory does not exist.
        IOError: If there is an issue with reading or writing the vocabulary files.
        ValueError: If the provided filename_prefix is not a string.
        Exception: Any other unexpected error that may occur during the execution of the method.
    """
    if not os.path.isdir(save_directory):
        logger.error(f"Vocabulary path ({save_directory}) should be a directory")
        return
    vocab_file = os.path.join(
        save_directory, (filename_prefix + "-" if filename_prefix else "") + VOCAB_FILES_NAMES["vocab_file"]
    )
    merge_file = os.path.join(
        save_directory, (filename_prefix + "-" if filename_prefix else "") + VOCAB_FILES_NAMES["merges_file"]
    )

    with open(vocab_file, "w", encoding="utf-8") as f:
        f.write(json.dumps(self.encoder, indent=2, sort_keys=True, ensure_ascii=False) + "\n")

    index = 0
    with open(merge_file, "w", encoding="utf-8") as writer:
        writer.write("#version: 0.2\n")
        for bpe_tokens, token_index in sorted(self.bpe_ranks.items(), key=lambda kv: kv[1]):
            if index != token_index:
                logger.warning(
                    f"Saving vocabulary to {merge_file}: BPE merge indices are not consecutive."
                    " Please check that the tokenizer is not corrupted!"
                )
                index = token_index
            writer.write(" ".join(bpe_tokens) + "\n")
            index += 1

    entity_vocab_file = os.path.join(
        save_directory, (filename_prefix + "-" if filename_prefix else "") + VOCAB_FILES_NAMES["entity_vocab_file"]
    )

    with open(entity_vocab_file, "w", encoding="utf-8") as f:
        f.write(json.dumps(self.entity_vocab, indent=2, sort_keys=True, ensure_ascii=False) + "\n")

    return vocab_file, merge_file, entity_vocab_file

mindnlp.transformers.models.luke.tokenization_luke.bytes_to_unicode() cached

Returns list of utf-8 byte and a mapping to unicode strings. We specifically avoids mapping to whitespace/control characters the bpe code barfs on.

The reversible bpe codes work on unicode strings. This means you need a large # of unicode characters in your vocab if you want to avoid UNKs. When you're at something like a 10B token dataset you end up needing around 5K for decent coverage. This is a significant percentage of your normal, say, 32K bpe vocab. To avoid that, we want lookup tables between utf-8 bytes and unicode strings.

Source code in mindnlp\transformers\models\luke\tokenization_luke.py 
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
@lru_cache()
# Copied from transformers.models.roberta.tokenization_roberta.bytes_to_unicode
def bytes_to_unicode():
    """
    Returns list of utf-8 byte and a mapping to unicode strings. We specifically avoids mapping to whitespace/control
    characters the bpe code barfs on.

    The reversible bpe codes work on unicode strings. This means you need a large # of unicode characters in your vocab
    if you want to avoid UNKs. When you're at something like a 10B token dataset you end up needing around 5K for
    decent coverage. This is a significant percentage of your normal, say, 32K bpe vocab. To avoid that, we want lookup
    tables between utf-8 bytes and unicode strings.
    """
    bs = (
        list(range(ord("!"), ord("~") + 1)) + list(range(ord("¡"), ord("¬") + 1)) + list(range(ord("®"), ord("ÿ") + 1))
    )
    cs = bs[:]
    n = 0
    for b in range(2**8):
        if b not in bs:
            bs.append(b)
            cs.append(2**8 + n)
            n += 1
    cs = [chr(n) for n in cs]
    return dict(zip(bs, cs))

mindnlp.transformers.models.luke.tokenization_luke.get_pairs(word)

Return set of symbol pairs in a word.

Word is represented as tuple of symbols (symbols being variable-length strings).

Source code in mindnlp\transformers\models\luke\tokenization_luke.py 
178
179
180
181
182
183
184
185
186
187
188
189
def get_pairs(word):
    """
    Return set of symbol pairs in a word.

    Word is represented as tuple of symbols (symbols being variable-length strings).
    """
    pairs = set()
    prev_char = word[0]
    for char in word[1:]:
        pairs.add((prev_char, char))
        prev_char = char
    return pairs