initial release

README.md

@@ -1,3 +1,39 @@
----
-license: mit
----
+---
+language:
+- "de"
+tags:
+- "german"
+- "token-classification"
+- "pos"
+- "dependency-parsing"
+datasets:
+- "universal_dependencies"
+license: "mit"
+pipeline_tag: "token-classification"
+---
+
+# bert-large-german-upos
+
+## Model Description
+
+This is a BERT model pre-trained with [UD_German-HDT](https://github.com/UniversalDependencies/UD_German-HDT) for POS-tagging and dependency-parsing, derived from [gbert-large](https://huggingface.co/deepset/gbert-large). Every word is tagged by [UPOS](https://universaldependencies.org/u/pos/) (Universal Part-Of-Speech).
+
+## How to Use
+
+```py
+from transformers import AutoTokenizer,AutoModelForTokenClassification
+tokenizer=AutoTokenizer.from_pretrained("KoichiYasuoka/bert-large-german-upos")
+model=AutoModelForTokenClassification.from_pretrained("KoichiYasuoka/bert-large-german-upos")
+```
+
+or
+
+```py
+import esupar
+nlp=esupar.load("KoichiYasuoka/bert-large-german-upos")
+```
+
+## See Also
+
+[esupar](https://github.com/KoichiYasuoka/esupar): Tokenizer POS-tagger and Dependency-parser with BERT/RoBERTa models
+

config.json

@@ -0,0 +1,275 @@
+{
+  "architectures": [
+    "BertForTokenClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "classifier_dropout": null,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 1024,
+  "id2label": {
+    "0": "ADJ",
+    "1": "ADJ+NOUN",
+    "2": "ADP",
+    "3": "ADP+DET",
+    "4": "ADV",
+    "5": "AUX",
+    "6": "B-ADJ",
+    "7": "B-ADP",
+    "8": "B-ADP+DET",
+    "9": "B-ADV",
+    "10": "B-AUX",
+    "11": "B-CCONJ",
+    "12": "B-DET",
+    "13": "B-INTJ",
+    "14": "B-NOUN",
+    "15": "B-NUM",
+    "16": "B-PRON",
+    "17": "B-PROPN",
+    "18": "B-PUNCT",
+    "19": "B-SCONJ",
+    "20": "B-VERB",
+    "21": "B-X",
+    "22": "CCONJ",
+    "23": "DET",
+    "24": "I-ADJ",
+    "25": "I-ADP",
+    "26": "I-ADP+DET",
+    "27": "I-ADV",
+    "28": "I-AUX",
+    "29": "I-CCONJ",
+    "30": "I-DET",
+    "31": "I-INTJ",
+    "32": "I-NOUN",
+    "33": "I-NUM",
+    "34": "I-PRON",
+    "35": "I-PROPN",
+    "36": "I-PUNCT",
+    "37": "I-SCONJ",
+    "38": "I-VERB",
+    "39": "I-X",
+    "40": "INTJ",
+    "41": "NOUN",
+    "42": "NUM",
+    "43": "PART",
+    "44": "PRON",
+    "45": "PROPN",
+    "46": "PUNCT",
+    "47": "SCONJ",
+    "48": "SYM",
+    "49": "VERB",
+    "50": "X"
+  },
+  "initializer_range": 0.02,
+  "intermediate_size": 4096,
+  "label2id": {
+    "ADJ": 0,
+    "ADJ+NOUN": 1,
+    "ADP": 2,
+    "ADP+DET": 3,
+    "ADV": 4,
+    "AUX": 5,
+    "B-ADJ": 6,
+    "B-ADP": 7,
+    "B-ADP+DET": 8,
+    "B-ADV": 9,
+    "B-AUX": 10,
+    "B-CCONJ": 11,
+    "B-DET": 12,
+    "B-INTJ": 13,
+    "B-NOUN": 14,
+    "B-NUM": 15,
+    "B-PRON": 16,
+    "B-PROPN": 17,
+    "B-PUNCT": 18,
+    "B-SCONJ": 19,
+    "B-VERB": 20,
+    "B-X": 21,
+    "CCONJ": 22,
+    "DET": 23,
+    "I-ADJ": 24,
+    "I-ADP": 25,
+    "I-ADP+DET": 26,
+    "I-ADV": 27,
+    "I-AUX": 28,
+    "I-CCONJ": 29,
+    "I-DET": 30,
+    "I-INTJ": 31,
+    "I-NOUN": 32,
+    "I-NUM": 33,
+    "I-PRON": 34,
+    "I-PROPN": 35,
+    "I-PUNCT": 36,
+    "I-SCONJ": 37,
+    "I-VERB": 38,
+    "I-X": 39,
+    "INTJ": 40,
+    "NOUN": 41,
+    "NUM": 42,
+    "PART": 43,
+    "PRON": 44,
+    "PROPN": 45,
+    "PUNCT": 46,
+    "SCONJ": 47,
+    "SYM": 48,
+    "VERB": 49,
+    "X": 50
+  },
+  "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",
+  "task_specific_params": {
+    "upos_multiword": {
+      "ADJ+NOUN": {
+        "selbst\u00e4ndigeDenken\u00b8": [
+          "selbst\u00e4ndige",
+          "Denken\u00b8"
+        ]
+      },
+      "ADP+DET": {
+        "A.": [
+          "AN",
+          "DEM"
+        ],
+        "Am": [
+          "An",
+          "dem"
+        ],
+        "Ans": [
+          "An",
+          "das"
+        ],
+        "Beim": [
+          "Bei",
+          "dem"
+        ],
+        "F\u00fcrs": [
+          "F\u00fcr",
+          "das"
+        ],
+        "Im": [
+          "In",
+          "dem"
+        ],
+        "Ins": [
+          "In",
+          "das"
+        ],
+        "Ums": [
+          "Um",
+          "das"
+        ],
+        "Unterm": [
+          "Unter",
+          "dem"
+        ],
+        "Vom": [
+          "Von",
+          "dem"
+        ],
+        "Zum": [
+          "Zu",
+          "dem"
+        ],
+        "Zur": [
+          "Zu",
+          "der"
+        ],
+        "am": [
+          "an",
+          "dem"
+        ],
+        "ans": [
+          "an",
+          "das"
+        ],
+        "aufs": [
+          "auf",
+          "das"
+        ],
+        "beim": [
+          "bei",
+          "dem"
+        ],
+        "durchs": [
+          "durch",
+          "das"
+        ],
+        "f\u00fcrs": [
+          "f\u00fcr",
+          "das"
+        ],
+        "hinterm": [
+          "hinter",
+          "dem"
+        ],
+        "hinters": [
+          "hinter",
+          "das"
+        ],
+        "im": [
+          "in",
+          "dem"
+        ],
+        "ins": [
+          "in",
+          "das"
+        ],
+        "ums": [
+          "um",
+          "das"
+        ],
+        "unterm": [
+          "unter",
+          "dem"
+        ],
+        "unters": [
+          "unter",
+          "das"
+        ],
+        "vom": [
+          "von",
+          "dem"
+        ],
+        "vorm": [
+          "vor",
+          "dem"
+        ],
+        "vors": [
+          "vor",
+          "das"
+        ],
+        "z.": [
+          "zu",
+          "dem"
+        ],
+        "zum": [
+          "zu",
+          "dem"
+        ],
+        "zur": [
+          "zu",
+          "der"
+        ],
+        "\u00dcbers": [
+          "\u00dcber",
+          "das"
+        ],
+        "\u00fcbers": [
+          "\u00fcber",
+          "das"
+        ]
+      }
+    }
+  },
+  "tokenizer_class": "BertTokenizerFast",
+  "torch_dtype": "float32",
+  "transformers_version": "4.14.1",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 31102
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5ecdea9d42e912d9c063b4d41d8b9303228459b4a2ecaa251b698a774b540618
+size 1339114865

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:c64f0a1fab8269b2ccf6aadb2b8fa773c9784b7a6009c773f90c9a0cdfe80750
+size 1420617381

tokenizer_config.json

@@ -0,0 +1 @@
+{"do_lower_case": false, "unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]", "tokenize_chinese_chars": true, "strip_accents": false, "model_max_length": 512, "do_basic_tokenize": true, "tokenizer_class": "BertTokenizerFast"}