File size: 8,511 Bytes
ab8d3d5
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
# coding=utf-8
# Copyright 2024 The HuggingFace Inc. team, The Hugging Face Team.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
"""Tokenization classes for FLMR."""


from typing import List, Optional, Union

from transformers.utils import TensorType, logging
from transformers.models.bert.tokenization_bert import BertTokenizer


logger = logging.get_logger(__name__)

VOCAB_FILES_NAMES = {"vocab_file": "vocab.txt", "tokenizer_file": "tokenizer_config.json"}

CONTEXT_ENCODER_PRETRAINED_VOCAB_FILES_MAP = {
    "vocab_file": {
        "LinWeizheDragon/PreFLMR_ViT-L": (
            "https://huggingface.co/LinWeizheDragon/PreFLMR_ViT-L/resolve/main/context_tokenizer/vocab.txt"
        ),
        "LinWeizheDragon/FLMR": (
            "https://huggingface.co/LinWeizheDragon/FLMR/resolve/main/context_tokenizer/vocab.txt"
        ),
    },
    "tokenizer_file": {
        "LinWeizheDragon/PreFLMR_ViT-L": (
            "https://huggingface.co/LinWeizheDragon/PreFLMR_ViT-L/resolve/main/context_tokenizer/tokenizer_config.json"
        ),
        "LinWeizheDragon/FLMR": (
            "https://huggingface.co/LinWeizheDragon/FLMR/resolve/main/context_tokenizer/tokenizer_config.json"
        ),
    },
}
QUESTION_ENCODER_PRETRAINED_VOCAB_FILES_MAP = {
    "vocab_file": {
        "LinWeizheDragon/PreFLMR_ViT-L": (
            "https://huggingface.co/LinWeizheDragon/PreFLMR_ViT-L/resolve/main/query_tokenizer/vocab.txt"
        ),
        "LinWeizheDragon/FLMR": ("https://huggingface.co/LinWeizheDragon/FLMR/resolve/main/query_tokenizer/vocab.txt"),
    },
    "tokenizer_file": {
        "LinWeizheDragon/PreFLMR_ViT-L": (
            "https://huggingface.co/LinWeizheDragon/PreFLMR_ViT-L/resolve/main/query_tokenizer/tokenizer_config.json"
        ),
        "LinWeizheDragon/FLMR": (
            "https://huggingface.co/LinWeizheDragon/FLMR/resolve/main/query_tokenizer/tokenizer_config.json"
        ),
    },
}


CONTEXT_ENCODER_PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES = {
    "LinWeizheDragon/PreFLMR_ViT-L": 512,
    "LinWeizheDragon/FLMR": 512,
}
QUESTION_ENCODER_PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES = {
    "LinWeizheDragon/PreFLMR_ViT-L": 512,
    "LinWeizheDragon/FLMR": 512,
}


CONTEXT_ENCODER_PRETRAINED_INIT_CONFIGURATION = {
    "LinWeizheDragon/PreFLMR_ViT-L": {"do_lower_case": True},
    "LinWeizheDragon/FLMR": {"do_lower_case": True},
}
QUESTION_ENCODER_PRETRAINED_INIT_CONFIGURATION = {
    "LinWeizheDragon/PreFLMR_ViT-L": {"do_lower_case": True},
    "LinWeizheDragon/FLMR": {"do_lower_case": True},
}


# Modified from colbert.modeling.tokenization
class FLMRContextEncoderTokenizer(BertTokenizer):
    r"""
    Construct a FLMRContextEncoder tokenizer.

    [`FLMRContextEncoderTokenizer`] is identical to [`BertTokenizer`] and runs end-to-end tokenization: punctuation
    splitting and wordpiece.

    Refer to superclass [`BertTokenizer`] for usage examples and documentation concerning parameters.
    """

    vocab_files_names = VOCAB_FILES_NAMES
    pretrained_vocab_files_map = CONTEXT_ENCODER_PRETRAINED_VOCAB_FILES_MAP
    max_model_input_sizes = CONTEXT_ENCODER_PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES
    pretrained_init_configuration = CONTEXT_ENCODER_PRETRAINED_INIT_CONFIGURATION

    def __init__(
        self,
        doc_maxlen: Optional[int] = 512,
        **kwargs,
    ):
        super().__init__(
            doc_maxlen=doc_maxlen,
            **kwargs,
        )

        self.doc_maxlen = doc_maxlen
        self.D_marker_token, self.D_marker_token_id = "[D]", self.convert_tokens_to_ids("[unused1]")

    def __call__(
        self,
        text: List[str],
        padding: Optional[Union[str, bool]] = "max_length",
        truncation: Optional[Union[bool, str]] = "longest_first",
        max_length: Optional[int] = 512,
        return_tensors: Optional[Union[str, TensorType]] = "pt",
        **kwargs,
    ):
        # add placehold for the [D] marker
        text = [". " + x for x in text]

        if max_length > self.doc_maxlen:
            # can not exceed the pre-set length
            max_length = self.doc_maxlen

        encoding = super().__call__(
            text,
            padding=padding,
            truncation=truncation,
            return_tensors=return_tensors,
            max_length=max_length,
            **kwargs,
        )

        ids, mask = encoding["input_ids"], encoding["attention_mask"]

        # postprocess for the [D] marker
        ids[:, 1] = self.D_marker_token_id

        # if bsize:
        #     # This bsize function is used in the original ColBERT codebase to split inputs into multiple batches
        #     if image_features is not None:
        #         ids, mask, image_features, reverse_indices = _sort_by_length(ids, mask, bsize, image_features=image_features)
        #         batches = _split_into_batches(ids, mask, bsize, image_features=image_features)
        #     else:
        #         ids, mask, reverse_indices = _sort_by_length(ids, mask, bsize)
        #         batches = _split_into_batches(ids, mask, bsize)

        #     return batches, reverse_indices

        encoding["input_ids"] = ids
        encoding["attention_mask"] = mask

        return encoding


# Modified from colbert.modeling.tokenization
class FLMRQueryEncoderTokenizer(BertTokenizer):
    r"""
    Constructs a FLMRQueryEncoder tokenizer.

    [`FLMRQueryEncoder`] is identical to [`BertTokenizer`] and runs end-to-end tokenization: punctuation
    splitting and wordpiece.

    Refer to superclass [`BertTokenizer`] for usage examples and documentation concerning parameters.
    """

    vocab_files_names = VOCAB_FILES_NAMES
    pretrained_vocab_files_map = QUESTION_ENCODER_PRETRAINED_VOCAB_FILES_MAP
    max_model_input_sizes = QUESTION_ENCODER_PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES
    pretrained_init_configuration = QUESTION_ENCODER_PRETRAINED_INIT_CONFIGURATION

    def __init__(
        self,
        *args,
        query_maxlen: Optional[int] = 32,
        attend_to_mask_tokens: Optional[bool] = False,
        **kwargs,
    ):
        super().__init__(
            *args,
            query_maxlen=query_maxlen,
            attend_to_mask_tokens=attend_to_mask_tokens,
            **kwargs,
        )

        self.query_maxlen = query_maxlen
        self.background_maxlen = 512 - self.query_maxlen + 1  # FIXME: Make this configurable
        self.attend_to_mask_tokens = attend_to_mask_tokens

        self.Q_marker_token, self.Q_marker_token_id = "[Q]", self.convert_tokens_to_ids("[unused0]")

    def __call__(
        self,
        text: Union[str, List[str]],
        padding: Optional[Union[str, bool]] = "max_length",
        truncation: Optional[Union[bool, str]] = True,
        max_length: Optional[int] = None,
        return_tensors: Optional[Union[str, TensorType]] = "pt",
        **kwargs,
    ):
        if isinstance(text, str):
            # convert to list if input is a single string
            text = [text]

        # add placehold for the [Q] marker
        text = [". " + x for x in text]

        if max_length is not None:
            # use user specified max_length
            pass
        else:
            # use default max length
            max_length = self.query_maxlen

        encoding = super().__call__(
            text,
            padding=padding,
            truncation=truncation,
            return_tensors=return_tensors,
            max_length=max_length,
            **kwargs,
        )

        ids, mask = encoding["input_ids"], encoding["attention_mask"]

        # postprocess for the [Q] marker and the [MASK] augmentation
        ids[:, 1] = self.Q_marker_token_id
        ids[ids == self.pad_token_id] = self.mask_token_id

        if self.attend_to_mask_tokens:
            # When attend_to_mask_tokens is True, we want to attend to the [MASK] tokens
            mask[ids == self.mask_token_id] = 1
            assert mask.sum().item() == mask.size(0) * mask.size(1), mask

        return {"input_ids": ids, "attention_mask": mask}