upload

README.md

@@ -1,5 +1,22 @@
----
-license: apache-2.0
----
+# GPLinker关系抽取模型
 
-GPLinker关系抽取模型
+## 模型介绍
+
++ 数据集：百度 `DUIE2.0` 信息抽取
+
++ 模型方法：[GPLinker：基于GlobalPointer的实体关系联合抽取](https://kexue.fm/archives/8888)
+
+## 使用方法
+
+```commandline
+pip install lightningnlp
+```
+
+```python
+from pprint import pprint
+from lightningnlp.task.relation_extraction import RelationExtractionPipeline
+
+pipline = RelationExtractionPipeline(model_name_or_path="xusenlin/duie-gplinker", model_name="gplinker", model_type="bert")
+text = "查尔斯·阿兰基斯（Charles Aránguiz），1989年4月17日出生于智利圣地亚哥，智利职业足球运动员，司职中场，效力于德国足球甲级联赛勒沃库森足球俱乐部。"
+pprint(pipline(text))
+```

config.json

@@ -0,0 +1,138 @@
+{
+  "_name_or_path": "hfl/chinese-roberta-wwm-ext",
+  "architectures": [
+    "GPLinker"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "bos_token_id": 0,
+  "classifier_dropout": null,
+  "directionality": "bidi",
+  "eos_token_id": 2,
+  "head_size": 64,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "id2predicate": {
+    "0": "\u4e08\u592b",
+    "1": "\u4e0a\u6620\u65f6\u95f4",
+    "2": "\u4e13\u4e1a\u4ee3\u7801",
+    "3": "\u4e3b\u6301\u4eba",
+    "4": "\u4e3b\u6f14",
+    "5": "\u4e3b\u89d2",
+    "6": "\u4eba\u53e3\u6570\u91cf",
+    "7": "\u4f5c\u66f2",
+    "8": "\u4f5c\u8005",
+    "9": "\u4f5c\u8bcd",
+    "10": "\u4fee\u4e1a\u5e74\u9650",
+    "11": "\u51fa\u54c1\u516c\u53f8",
+    "12": "\u51fa\u7248\u793e",
+    "13": "\u51fa\u751f\u5730",
+    "14": "\u51fa\u751f\u65e5\u671f",
+    "15": "\u521b\u59cb\u4eba",
+    "16": "\u5236\u7247\u4eba",
+    "17": "\u5360\u5730\u9762\u79ef",
+    "18": "\u53f7",
+    "19": "\u5609\u5bbe",
+    "20": "\u56fd\u7c4d",
+    "21": "\u59bb\u5b50",
+    "22": "\u5b57",
+    "23": "\u5b98\u65b9\u8bed\u8a00",
+    "24": "\u5bfc\u6f14",
+    "25": "\u603b\u90e8\u5730\u70b9",
+    "26": "\u6210\u7acb\u65e5\u671f",
+    "27": "\u6240\u5728\u57ce\u5e02",
+    "28": "\u6240\u5c5e\u4e13\u8f91",
+    "29": "\u6539\u7f16\u81ea",
+    "30": "\u671d\u4ee3",
+    "31": "\u6b4c\u624b",
+    "32": "\u6bcd\u4eb2",
+    "33": "\u6bd5\u4e1a\u9662\u6821",
+    "34": "\u6c11\u65cf",
+    "35": "\u6c14\u5019",
+    "36": "\u6ce8\u518c\u8d44\u672c",
+    "37": "\u6d77\u62d4",
+    "38": "\u7236\u4eb2",
+    "39": "\u76ee",
+    "40": "\u7956\u7c4d",
+    "41": "\u7b80\u79f0",
+    "42": "\u7f16\u5267",
+    "43": "\u8463\u4e8b\u957f",
+    "44": "\u8eab\u9ad8",
+    "45": "\u8fde\u8f7d\u7f51\u7ad9",
+    "46": "\u90ae\u653f\u7f16\u7801",
+    "47": "\u9762\u79ef",
+    "48": "\u9996\u90fd"
+  },
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "bert",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "num_predicates": 49,
+  "output_past": true,
+  "pad_token_id": 0,
+  "pooler_fc_size": 768,
+  "pooler_num_attention_heads": 12,
+  "pooler_num_fc_layers": 3,
+  "pooler_size_per_head": 128,
+  "pooler_type": "first_token_transform",
+  "position_embedding_type": "absolute",
+  "predicate2id": {
+    "\u4e08\u592b": 0,
+    "\u4e0a\u6620\u65f6\u95f4": 1,
+    "\u4e13\u4e1a\u4ee3\u7801": 2,
+    "\u4e3b\u6301\u4eba": 3,
+    "\u4e3b\u6f14": 4,
+    "\u4e3b\u89d2": 5,
+    "\u4eba\u53e3\u6570\u91cf": 6,
+    "\u4f5c\u66f2": 7,
+    "\u4f5c\u8005": 8,
+    "\u4f5c\u8bcd": 9,
+    "\u4fee\u4e1a\u5e74\u9650": 10,
+    "\u51fa\u54c1\u516c\u53f8": 11,
+    "\u51fa\u7248\u793e": 12,
+    "\u51fa\u751f\u5730": 13,
+    "\u51fa\u751f\u65e5\u671f": 14,
+    "\u521b\u59cb\u4eba": 15,
+    "\u5236\u7247\u4eba": 16,
+    "\u5360\u5730\u9762\u79ef": 17,
+    "\u53f7": 18,
+    "\u5609\u5bbe": 19,
+    "\u56fd\u7c4d": 20,
+    "\u59bb\u5b50": 21,
+    "\u5b57": 22,
+    "\u5b98\u65b9\u8bed\u8a00": 23,
+    "\u5bfc\u6f14": 24,
+    "\u603b\u90e8\u5730\u70b9": 25,
+    "\u6210\u7acb\u65e5\u671f": 26,
+    "\u6240\u5728\u57ce\u5e02": 27,
+    "\u6240\u5c5e\u4e13\u8f91": 28,
+    "\u6539\u7f16\u81ea": 29,
+    "\u671d\u4ee3": 30,
+    "\u6b4c\u624b": 31,
+    "\u6bcd\u4eb2": 32,
+    "\u6bd5\u4e1a\u9662\u6821": 33,
+    "\u6c11\u65cf": 34,
+    "\u6c14\u5019": 35,
+    "\u6ce8\u518c\u8d44\u672c": 36,
+    "\u6d77\u62d4": 37,
+    "\u7236\u4eb2": 38,
+    "\u76ee": 39,
+    "\u7956\u7c4d": 40,
+    "\u7b80\u79f0": 41,
+    "\u7f16\u5267": 42,
+    "\u8463\u4e8b\u957f": 43,
+    "\u8eab\u9ad8": 44,
+    "\u8fde\u8f7d\u7f51\u7ad9": 45,
+    "\u90ae\u653f\u7f16\u7801": 46,
+    "\u9762\u79ef": 47,
+    "\u9996\u90fd": 48
+  },
+  "torch_dtype": "float32",
+  "transformers_version": "4.22.1",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 21128
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5875be88fd45bc4585ad0cc60f12727749df74a62f8bf383a9e7e1978bc5c920
+size 408328921

special_tokens_map.json

@@ -0,0 +1,7 @@
+{
+  "cls_token": "[CLS]",
+  "mask_token": "[MASK]",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "unk_token": "[UNK]"
+}

tokenizer_config.json

@@ -0,0 +1,13 @@
+{
+  "cls_token": "[CLS]",
+  "do_lower_case": true,
+  "mask_token": "[MASK]",
+  "name_or_path": "hfl/chinese-roberta-wwm-ext",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "special_tokens_map_file": "/home/xusenlin/.cache/huggingface/hub/models--hfl--chinese-roberta-wwm-ext/snapshots/5c58d0b8ec1d9014354d691c538661bf00bfdb44/special_tokens_map.json",
+  "strip_accents": null,
+  "tokenize_chinese_chars": true,
+  "tokenizer_class": "BertTokenizer",
+  "unk_token": "[UNK]"
+}