initial release

README.md

@@ -1,3 +1,54 @@
----
-license: cc-by-sa-4.0
----
+---
+language:
+- "ja"
+tags:
+- "japanese"
+- "token-classification"
+- "pos"
+- "wikipedia"
+- "dependency-parsing"
+datasets:
+- "universal_dependencies"
+license: "cc-by-sa-4.0"
+pipeline_tag: "token-classification"
+widget:
+- text: "国境の長いトンネルを抜けると雪国であった。"
+---
+
+# bert-large-japanese-unidic-luw-upos
+
+## Model Description
+
+This is a BERT model pre-trained on Japanese Wikipedia texts for POS-tagging and dependency-parsing, derived from [bert-large-japanese](https://huggingface.co/cl-tohoku/bert-large-japanese). Every long-unit-word is tagged by [UPOS](https://universaldependencies.org/u/pos/) (Universal Part-Of-Speech).
+
+## How to Use
+
+```py
+import torch
+from transformers import AutoTokenizer,AutoModelForTokenClassification
+tokenizer=AutoTokenizer.from_pretrained("KoichiYasuoka/bert-large-japanese-unidic-luw-upos")
+model=AutoModelForTokenClassification.from_pretrained("KoichiYasuoka/bert-large-japanese-unidic-luw-upos")
+s="国境の長いトンネルを抜けると雪国であった。"
+t=tokenizer.tokenize(s)
+p=[model.config.id2label[q] for q in torch.argmax(model(tokenizer.encode(s,return_tensors="pt"))["logits"],dim=2)[0].tolist()[1:-1]]
+print(list(zip(t,p)))
+```
+
+or
+
+```py
+import esupar
+nlp=esupar.load("KoichiYasuoka/bert-large-japanese-unidic-luw-upos")
+print(nlp("国境の長いトンネルを抜けると雪国であった。"))
+```
+
+[fugashi](https://pypi.org/project/fugashi), [unidic-lite](https://pypi.org/project/unidic-lite) and [pytokenizations](https://pypi.org/project/pytokenizations) are required.
+
+## Reference
+
+安岡孝一: [Transformersと国語研長単位による日本語係り受け解析モデルの製作](http://id.nii.ac.jp/1001/00216223/), 情報処理学会研究報告, Vol.2022-CH-128, No.7 (2022年2月), pp.1-8.
+
+## See Also
+
+[esupar](https://github.com/KoichiYasuoka/esupar): Tokenizer POS-tagger and Dependency-parser with BERT/RoBERTa models
+

config.json

@@ -0,0 +1,131 @@
+{
+  "architectures": [
+    "BertForTokenClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "classifier_dropout": null,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 1024,
+  "id2label": {
+    "0": "B-SYM",
+    "1": "B-INTJ",
+    "2": "I-ADP",
+    "3": "I-X",
+    "4": "ADP",
+    "5": "PUNCT",
+    "6": "B-VERB",
+    "7": "I-VERB",
+    "8": "I-NUM",
+    "9": "VERB",
+    "10": "PRON",
+    "11": "I-DET",
+    "12": "B-ADP",
+    "13": "PROPN",
+    "14": "I-PUNCT",
+    "15": "I-CCONJ",
+    "16": "NUM",
+    "17": "I-INTJ",
+    "18": "AUX",
+    "19": "INTJ",
+    "20": "CCONJ",
+    "21": "I-PRON",
+    "22": "B-CCONJ",
+    "23": "X",
+    "24": "B-PUNCT",
+    "25": "I-SYM",
+    "26": "I-SCONJ",
+    "27": "SCONJ",
+    "28": "NOUN",
+    "29": "DET",
+    "30": "ADV",
+    "31": "PART",
+    "32": "B-PRON",
+    "33": "I-AUX",
+    "34": "B-NUM",
+    "35": "I-ADJ",
+    "36": "B-SCONJ",
+    "37": "I-PART",
+    "38": "I-NOUN",
+    "39": "I-ADV",
+    "40": "ADJ",
+    "41": "B-X",
+    "42": "B-AUX",
+    "43": "B-PROPN",
+    "44": "B-DET",
+    "45": "B-ADV",
+    "46": "I-PROPN",
+    "47": "B-NOUN",
+    "48": "SYM",
+    "49": "B-PART",
+    "50": "B-ADJ"
+  },
+  "initializer_range": 0.02,
+  "intermediate_size": 4096,
+  "label2id": {
+    "ADJ": 40,
+    "ADP": 4,
+    "ADV": 30,
+    "AUX": 18,
+    "B-ADJ": 50,
+    "B-ADP": 12,
+    "B-ADV": 45,
+    "B-AUX": 42,
+    "B-CCONJ": 22,
+    "B-DET": 44,
+    "B-INTJ": 1,
+    "B-NOUN": 47,
+    "B-NUM": 34,
+    "B-PART": 49,
+    "B-PRON": 32,
+    "B-PROPN": 43,
+    "B-PUNCT": 24,
+    "B-SCONJ": 36,
+    "B-SYM": 0,
+    "B-VERB": 6,
+    "B-X": 41,
+    "CCONJ": 20,
+    "DET": 29,
+    "I-ADJ": 35,
+    "I-ADP": 2,
+    "I-ADV": 39,
+    "I-AUX": 33,
+    "I-CCONJ": 15,
+    "I-DET": 11,
+    "I-INTJ": 17,
+    "I-NOUN": 38,
+    "I-NUM": 8,
+    "I-PART": 37,
+    "I-PRON": 21,
+    "I-PROPN": 46,
+    "I-PUNCT": 14,
+    "I-SCONJ": 26,
+    "I-SYM": 25,
+    "I-VERB": 7,
+    "I-X": 3,
+    "INTJ": 19,
+    "NOUN": 28,
+    "NUM": 16,
+    "PART": 31,
+    "PRON": 10,
+    "PROPN": 13,
+    "PUNCT": 5,
+    "SCONJ": 27,
+    "SYM": 48,
+    "VERB": 9,
+    "X": 23
+  },
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "bert",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 24,
+  "pad_token_id": 0,
+  "position_embedding_type": "absolute",
+  "tokenizer_class": "BertJapaneseTokenizer",
+  "torch_dtype": "float32",
+  "transformers_version": "4.11.3",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 32768
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c26b02e8926191ac7f7fb3340d3d286e08fe213b8b34a0dafe120ad0b2f00bec
+size 1345944210

special_tokens_map.json

@@ -0,0 +1 @@
+{"unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]"}

supar.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b2e1383c4cbb6c4a5f2a2a09cb8a89e726415291e030194772840dbcc68f68d9
+size 1399761610

tokenizer_config.json

@@ -0,0 +1 @@
+{"unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]", "do_lower_case": false, "do_word_tokenize": true, "do_subword_tokenize": true, "word_tokenizer_type": "mecab", "subword_tokenizer_type": "wordpiece", "never_split": null, "mecab_kwargs": {"mecab_dic": "unidic_lite"}, "model_max_length": 512, "tokenizer_class": "BertJapaneseTokenizer"}