Add model files

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+entity_vocab.json filter=lfs diff=lfs merge=lfs -text

added_tokens.json

@@ -0,0 +1,4 @@
+{
+  "<ent2>": 32769,
+  "<ent>": 32768
+}

config.json

@@ -0,0 +1,32 @@
+{
+  "_name_or_path": "cl-tohoku/bert-base-japanese-v3",
+  "architectures": [
+    "LukeForMaskedLM"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "bert_model_name": "cl-tohoku/bert-base-japanese-v3",
+  "bos_token_id": null,
+  "classifier_dropout": null,
+  "cls_entity_prediction": false,
+  "entity_emb_size": 256,
+  "entity_vocab_size": 591699,
+  "eos_token_id": null,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "luke",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "pad_token_id": 0,
+  "position_embedding_type": "absolute",
+  "torch_dtype": "float32",
+  "transformers_version": "4.30.2",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "use_entity_aware_attention": true,
+  "vocab_size": 32770
+}

entity_vocab.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:be6327e7cafc2f2b5f694a594d57113fd2bf6b620c592929202f75683b18b67d
+size 23721849

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:12fc608cd4f1662905c6e025fea20ca90f8494fa93a5c1f7c825ed41220ef2e7
+size 1143901513

special_tokens_map.json

@@ -0,0 +1,23 @@
+{
+  "additional_special_tokens": [
+    {
+      "content": "<ent>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false
+    },
+    {
+      "content": "<ent2>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false
+    }
+  ],
+  "cls_token": "[CLS]",
+  "mask_token": "[MASK]",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "unk_token": "[UNK]"
+}

tokenization_luke_bert_japanese.py

@@ -0,0 +1,420 @@
+# coding=utf-8
+# Copyright Studio-Ouisa and The HuggingFace Inc. team. All rights reserved.
+#
+# 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 LUKE."""
+
+import collections
+import copy
+import json
+import os
+from typing import List, Optional, Tuple
+
+from transformers.models.bert_japanese.tokenization_bert_japanese import (
+    BasicTokenizer,
+    CharacterTokenizer,
+    JumanppTokenizer,
+    MecabTokenizer,
+    SentencepieceTokenizer,
+    SudachiTokenizer,
+    WordpieceTokenizer,
+    load_vocab,
+)
+from transformers.models.luke import LukeTokenizer
+from transformers.tokenization_utils_base import AddedToken
+from transformers.utils import logging
+
+
+logger = logging.get_logger(__name__)
+
+EntitySpan = Tuple[int, int]
+EntitySpanInput = List[EntitySpan]
+Entity = str
+EntityInput = List[Entity]
+
+VOCAB_FILES_NAMES = {"vocab_file": "vocab.txt", "entity_vocab_file": "entity_vocab.json"}
+
+PRETRAINED_VOCAB_FILES_MAP = {"vocab_file": {}, "entity_vocab_file": {}}
+
+PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES = {}
+
+
+class LukeBertJapaneseTokenizer(LukeTokenizer):
+    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,
+        entity_vocab_file,
+        spm_file=None,
+        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]",
+        do_lower_case=False,
+        do_word_tokenize=True,
+        do_subword_tokenize=True,
+        word_tokenizer_type="basic",
+        subword_tokenizer_type="wordpiece",
+        never_split=None,
+        unk_token="[UNK]",
+        sep_token="[SEP]",
+        pad_token="[PAD]",
+        cls_token="[CLS]",
+        mask_token="[MASK]",
+        mecab_kwargs=None,
+        sudachi_kwargs=None,
+        jumanpp_kwargs=None,
+        **kwargs,
+    ):
+        # We call the grandparent's init, not the parent's.
+        super(LukeTokenizer, self).__init__(
+            spm_file=spm_file,
+            unk_token=unk_token,
+            sep_token=sep_token,
+            pad_token=pad_token,
+            cls_token=cls_token,
+            mask_token=mask_token,
+            do_lower_case=do_lower_case,
+            do_word_tokenize=do_word_tokenize,
+            do_subword_tokenize=do_subword_tokenize,
+            word_tokenizer_type=word_tokenizer_type,
+            subword_tokenizer_type=subword_tokenizer_type,
+            never_split=never_split,
+            mecab_kwargs=mecab_kwargs,
+            sudachi_kwargs=sudachi_kwargs,
+            jumanpp_kwargs=jumanpp_kwargs,
+            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,
+        )
+
+        if subword_tokenizer_type == "sentencepiece":
+            if not os.path.isfile(spm_file):
+                raise ValueError(
+                    f"Can't find a vocabulary file at path '{spm_file}'. To load the vocabulary from a Google"
+                    " pretrained model use `tokenizer = AutoTokenizer.from_pretrained(PRETRAINED_MODEL_NAME)`"
+                )
+            self.spm_file = spm_file
+        else:
+            if not os.path.isfile(vocab_file):
+                raise ValueError(
+                    f"Can't find a vocabulary file at path '{vocab_file}'. To load the vocabulary from a Google"
+                    " pretrained model use `tokenizer = AutoTokenizer.from_pretrained(PRETRAINED_MODEL_NAME)`"
+                )
+            self.vocab = load_vocab(vocab_file)
+            self.ids_to_tokens = collections.OrderedDict([(ids, tok) for tok, ids in self.vocab.items()])
+
+        self.do_word_tokenize = do_word_tokenize
+        self.word_tokenizer_type = word_tokenizer_type
+        self.lower_case = do_lower_case
+        self.never_split = never_split
+        self.mecab_kwargs = copy.deepcopy(mecab_kwargs)
+        self.sudachi_kwargs = copy.deepcopy(sudachi_kwargs)
+        self.jumanpp_kwargs = copy.deepcopy(jumanpp_kwargs)
+        if do_word_tokenize:
+            if word_tokenizer_type == "basic":
+                self.word_tokenizer = BasicTokenizer(
+                    do_lower_case=do_lower_case, never_split=never_split, tokenize_chinese_chars=False
+                )
+            elif word_tokenizer_type == "mecab":
+                self.word_tokenizer = MecabTokenizer(
+                    do_lower_case=do_lower_case, never_split=never_split, **(mecab_kwargs or {})
+                )
+            elif word_tokenizer_type == "sudachi":
+                self.word_tokenizer = SudachiTokenizer(
+                    do_lower_case=do_lower_case, never_split=never_split, **(sudachi_kwargs or {})
+                )
+            elif word_tokenizer_type == "jumanpp":
+                self.word_tokenizer = JumanppTokenizer(
+                    do_lower_case=do_lower_case, never_split=never_split, **(jumanpp_kwargs or {})
+                )
+            else:
+                raise ValueError(f"Invalid word_tokenizer_type '{word_tokenizer_type}' is specified.")
+
+        self.do_subword_tokenize = do_subword_tokenize
+        self.subword_tokenizer_type = subword_tokenizer_type
+        if do_subword_tokenize:
+            if subword_tokenizer_type == "wordpiece":
+                self.subword_tokenizer = WordpieceTokenizer(vocab=self.vocab, unk_token=self.unk_token)
+            elif subword_tokenizer_type == "character":
+                self.subword_tokenizer = CharacterTokenizer(vocab=self.vocab, unk_token=self.unk_token)
+            elif subword_tokenizer_type == "sentencepiece":
+                self.subword_tokenizer = SentencepieceTokenizer(vocab=self.spm_file, unk_token=self.unk_token)
+            else:
+                raise ValueError(f"Invalid subword_tokenizer_type '{subword_tokenizer_type}' is specified.")
+
+        # 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
+
+    @property
+    # Copied from BertJapaneseTokenizer
+    def do_lower_case(self):
+        return self.lower_case
+
+    # Copied from BertJapaneseTokenizer
+    def __getstate__(self):
+        state = dict(self.__dict__)
+        if self.word_tokenizer_type in ["mecab", "sudachi", "jumanpp"]:
+            del state["word_tokenizer"]
+        return state
+
+    # Copied from BertJapaneseTokenizer
+    def __setstate__(self, state):
+        self.__dict__ = state
+        if self.word_tokenizer_type == "mecab":
+            self.word_tokenizer = MecabTokenizer(
+                do_lower_case=self.do_lower_case, never_split=self.never_split, **(self.mecab_kwargs or {})
+            )
+        elif self.word_tokenizer_type == "sudachi":
+            self.word_tokenizer = SudachiTokenizer(
+                do_lower_case=self.do_lower_case, never_split=self.never_split, **(self.sudachi_kwargs or {})
+            )
+        elif self.word_tokenizer_type == "jumanpp":
+            self.word_tokenizer = JumanppTokenizer(
+                do_lower_case=self.do_lower_case, never_split=self.never_split, **(self.jumanpp_kwargs or {})
+            )
+
+    # Copied from BertJapaneseTokenizer
+    def _tokenize(self, text):
+        if self.do_word_tokenize:
+            tokens = self.word_tokenizer.tokenize(text, never_split=self.all_special_tokens)
+        else:
+            tokens = [text]
+
+        if self.do_subword_tokenize:
+            split_tokens = [sub_token for token in tokens for sub_token in self.subword_tokenizer.tokenize(token)]
+        else:
+            split_tokens = tokens
+
+        return split_tokens
+
+    @property
+    # Copied from BertJapaneseTokenizer
+    def vocab_size(self):
+        if self.subword_tokenizer_type == "sentencepiece":
+            return len(self.subword_tokenizer.sp_model)
+        return len(self.vocab)
+
+    # Copied from BertJapaneseTokenizer
+    def get_vocab(self):
+        if self.subword_tokenizer_type == "sentencepiece":
+            vocab = {self.convert_ids_to_tokens(i): i for i in range(self.vocab_size)}
+            vocab.update(self.added_tokens_encoder)
+            return vocab
+        return dict(self.vocab, **self.added_tokens_encoder)
+
+    # Copied from BertJapaneseTokenizer
+    def _convert_token_to_id(self, token):
+        """Converts a token (str) in an id using the vocab."""
+        if self.subword_tokenizer_type == "sentencepiece":
+            return self.subword_tokenizer.sp_model.PieceToId(token)
+        return self.vocab.get(token, self.vocab.get(self.unk_token))
+
+    # Copied from BertJapaneseTokenizer
+    def _convert_id_to_token(self, index):
+        """Converts an index (integer) in a token (str) using the vocab."""
+        if self.subword_tokenizer_type == "sentencepiece":
+            return self.subword_tokenizer.sp_model.IdToPiece(index)
+        return self.ids_to_tokens.get(index, self.unk_token)
+
+    # Copied from BertJapaneseTokenizer
+    def convert_tokens_to_string(self, tokens):
+        """Converts a sequence of tokens (string) in a single string."""
+        if self.subword_tokenizer_type == "sentencepiece":
+            return self.subword_tokenizer.sp_model.decode(tokens)
+        out_string = " ".join(tokens).replace(" ##", "").strip()
+        return out_string
+
+    # Copied from BertJapaneseTokenizer
+    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 BERT sequence has the following format:
+
+        - single sequence: `[CLS] X [SEP]`
+        - pair of sequences: `[CLS] A [SEP] B [SEP]`
+
+        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 + token_ids_1 + sep
+
+    # Copied from BertJapaneseTokenizer
+    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 not None:
+            return [1] + ([0] * len(token_ids_0)) + [1] + ([0] * len(token_ids_1)) + [1]
+        return [1] + ([0] * len(token_ids_0)) + [1]
+
+    # Copied from BertJapaneseTokenizer
+    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. A BERT sequence
+        pair mask has the following format:
+
+        ```
+        0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1
+        | first sequence    | second sequence |
+        ```
+
+        If `token_ids_1` is `None`, this method only returns the first portion of the mask (0s).
+
+        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 [token type IDs](../glossary#token-type-ids) according to the given sequence(s).
+        """
+        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) * [0] + len(token_ids_1 + sep) * [1]
+
+    def prepare_for_tokenization(self, text, is_split_into_words=False, **kwargs):
+        return (text, kwargs)
+
+    def save_vocabulary(self, save_directory: str, filename_prefix: Optional[str] = None) -> Tuple[str]:
+        if os.path.isdir(save_directory):
+            if self.subword_tokenizer_type == "sentencepiece":
+                vocab_file = os.path.join(
+                    save_directory, (filename_prefix + "-" if filename_prefix else "") + VOCAB_FILES_NAMES["spm_file"]
+                )
+            else:
+                vocab_file = os.path.join(
+                    save_directory,
+                    (filename_prefix + "-" if filename_prefix else "") + VOCAB_FILES_NAMES["vocab_file"],
+                )
+        else:
+            vocab_file = (filename_prefix + "-" if filename_prefix else "") + save_directory
+
+        if self.subword_tokenizer_type == "sentencepiece":
+            with open(vocab_file, "wb") as writer:
+                content_spiece_model = self.subword_tokenizer.sp_model.serialized_model_proto()
+                writer.write(content_spiece_model)
+        else:
+            with open(vocab_file, "w", encoding="utf-8") as writer:
+                index = 0
+                for token, token_index in sorted(self.vocab.items(), key=lambda kv: kv[1]):
+                    if index != token_index:
+                        logger.warning(
+                            f"Saving vocabulary to {vocab_file}: vocabulary indices are not consecutive."
+                            " Please check that the vocabulary is not corrupted!"
+                        )
+                        index = token_index
+                    writer.write(token + "\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, entity_vocab_file

tokenizer_config.json

@@ -0,0 +1,37 @@
+{
+  "auto_map": {
+    "AutoTokenizer": [
+      "tokenization_luke_bert_japanese.LukeBertJapaneseTokenizer",
+      null
+    ]
+  },
+  "clean_up_tokenization_spaces": true,
+  "cls_token": "[CLS]",
+  "do_lower_case": false,
+  "do_subword_tokenize": true,
+  "do_word_tokenize": true,
+  "entity_mask2_token": "[MASK2]",
+  "entity_mask_token": "[MASK]",
+  "entity_pad_token": "[PAD]",
+  "entity_token_1": "<ent>",
+  "entity_token_2": "<ent2>",
+  "entity_unk_token": "[UNK]",
+  "jumanpp_kwargs": null,
+  "mask_token": "[MASK]",
+  "max_entity_length": 32,
+  "max_mention_length": 30,
+  "mecab_kwargs": {
+    "mecab_dic": "unidic_lite"
+  },
+  "model_max_length": 512,
+  "never_split": null,
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "spm_file": null,
+  "subword_tokenizer_type": "wordpiece",
+  "sudachi_kwargs": null,
+  "task": null,
+  "tokenizer_class": "LukeBertJapaneseTokenizer",
+  "unk_token": "[UNK]",
+  "word_tokenizer_type": "mecab"
+}