Upload 6 files

Add 1st version

config.json

@@ -0,0 +1,148 @@
+{
+  "architectures": [
+    "BertForTokenClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "classifier_dropout": null,
+  "directionality": "bidi",
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "id2label": {
+    "0": "0",
+    "1": "B-H\u1ecd v\u00e0 t\u00ean",
+    "2": "I-H\u1ecd v\u00e0 t\u00ean",
+    "3": "B-Gi\u1edbi t\u00ednh",
+    "4": "I-Gi\u1edbi t\u00ednh",
+    "5": "B-S\u1ed1 \u0111i\u1ec7n tho\u1ea1i",
+    "6": "I-S\u1ed1 \u0111i\u1ec7n tho\u1ea1i",
+    "7": "B-Email",
+    "8": "I-Email",
+    "9": "B-\u0110\u1ecba ch\u1ec9",
+    "10": "I-\u0110\u1ecba ch\u1ec9",
+    "11": "B-N\u01a1i l\u00e0m vi\u1ec7c",
+    "12": "I-N\u01a1i l\u00e0m vi\u1ec7c",
+    "13": "B-Ngh\u1ec7 nghi\u1ec7p",
+    "14": "I-Ngh\u1ec7 nghi\u1ec7p",
+    "15": "B-D\u00e2n t\u1ed9c",
+    "16": "I-D\u00e2n t\u1ed9c",
+    "17": "B-T\u00f4n gi\u00e1o",
+    "18": "I-T\u00f4n gi\u00e1o",
+    "19": "B-N\u01a1i sinh",
+    "20": "I-N\u01a1i sinh",
+    "21": "B-Qu\u00ea qu\u00e1n",
+    "22": "I-Qu\u00ea qu\u00e1n",
+    "23": "B-Qu\u1ed1c t\u1ecbnh",
+    "24": "I-Qu\u1ed1c t\u1ecbnh",
+    "25": "B-Nh\u00f3m m\u00e1u",
+    "26": "I-Nh\u00f3m m\u00e1u",
+    "27": "B-S\u1ed1 c\u0103n c\u01b0\u1edbc c\u00f4ng d\u00e2n",
+    "28": "I-S\u1ed1 c\u0103n c\u01b0\u1edbc c\u00f4ng d\u00e2n",
+    "29": "B-S\u1ed1 h\u1ed9 chi\u1ebfu",
+    "30": "I-S\u1ed1 h\u1ed9 chi\u1ebfu",
+    "31": "B-N\u01a1i \u1edf hi\u1ec7n t\u1ea1i",
+    "32": "I-N\u01a1i \u1edf hi\u1ec7n t\u1ea1i",
+    "33": "B-T\u00ecnh tr\u1ea1ng h\u00f4n nh\u00e2n",
+    "34": "I-T\u00ecnh tr\u1ea1ng h\u00f4n nh\u00e2n",
+    "35": "B-Ng\u00e0y sinh",
+    "36": "I-Ng\u00e0y sinh",
+    "37": "B-Ng\u00e0nh h\u1ecdc",
+    "38": "I-Ng\u00e0nh h\u1ecdc",
+    "39": "B-Tr\u01b0\u1eddng h\u1ecdc",
+    "40": "I-Tr\u01b0\u1eddng h\u1ecdc",
+    "41": "B-B\u1eb1ng c\u1ea5p",
+    "42": "I-B\u1eb1ng c\u1ea5p",
+    "43": "B-Ch\u1ee9c v\u1ee5",
+    "44": "I-Ch\u1ee9c v\u1ee5",
+    "45": "B-S\u1edf th\u00edch",
+    "46": "I-S\u1edf th\u00edch",
+    "47": "B-T\u00ednh c\u00e1ch",
+    "48": "I-T\u00ednh c\u00e1ch",
+    "49": "B-\u0110i\u1ec3m m\u1ea1nh",
+    "50": "I-\u0110i\u1ec3m m\u1ea1nh",
+    "51": "B-\u0110i\u1ec3m y\u1ebfu",
+    "52": "I-\u0110i\u1ec3m y\u1ebfu",
+    "53": "B-Tu\u1ed5i",
+    "54": "I-Tu\u1ed5i",
+    "55": "B-M\u1ee5c ti\u00eau, mong mu\u1ed1n",
+    "56": "I-M\u1ee5c ti\u00eau, mong mu\u1ed1n"
+  },
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "label2id": {
+    "0": 0,
+    "B-B\u1eb1ng c\u1ea5p": 41,
+    "B-Ch\u1ee9c v\u1ee5": 43,
+    "B-D\u00e2n t\u1ed9c": 15,
+    "B-Email": 7,
+    "B-Gi\u1edbi t\u00ednh": 3,
+    "B-H\u1ecd v\u00e0 t\u00ean": 1,
+    "B-M\u1ee5c ti\u00eau, mong mu\u1ed1n": 55,
+    "B-Ngh\u1ec7 nghi\u1ec7p": 13,
+    "B-Ng\u00e0nh h\u1ecdc": 37,
+    "B-Ng\u00e0y sinh": 35,
+    "B-Nh\u00f3m m\u00e1u": 25,
+    "B-N\u01a1i l\u00e0m vi\u1ec7c": 11,
+    "B-N\u01a1i sinh": 19,
+    "B-N\u01a1i \u1edf hi\u1ec7n t\u1ea1i": 31,
+    "B-Qu\u00ea qu\u00e1n": 21,
+    "B-Qu\u1ed1c t\u1ecbnh": 23,
+    "B-S\u1ed1 c\u0103n c\u01b0\u1edbc c\u00f4ng d\u00e2n": 27,
+    "B-S\u1ed1 h\u1ed9 chi\u1ebfu": 29,
+    "B-S\u1ed1 \u0111i\u1ec7n tho\u1ea1i": 5,
+    "B-S\u1edf th\u00edch": 45,
+    "B-Tr\u01b0\u1eddng h\u1ecdc": 39,
+    "B-Tu\u1ed5i": 53,
+    "B-T\u00ecnh tr\u1ea1ng h\u00f4n nh\u00e2n": 33,
+    "B-T\u00ednh c\u00e1ch": 47,
+    "B-T\u00f4n gi\u00e1o": 17,
+    "B-\u0110i\u1ec3m m\u1ea1nh": 49,
+    "B-\u0110i\u1ec3m y\u1ebfu": 51,
+    "B-\u0110\u1ecba ch\u1ec9": 9,
+    "I-B\u1eb1ng c\u1ea5p": 42,
+    "I-Ch\u1ee9c v\u1ee5": 44,
+    "I-D\u00e2n t\u1ed9c": 16,
+    "I-Email": 8,
+    "I-Gi\u1edbi t\u00ednh": 4,
+    "I-H\u1ecd v\u00e0 t\u00ean": 2,
+    "I-M\u1ee5c ti\u00eau, mong mu\u1ed1n": 56,
+    "I-Ngh\u1ec7 nghi\u1ec7p": 14,
+    "I-Ng\u00e0nh h\u1ecdc": 38,
+    "I-Ng\u00e0y sinh": 36,
+    "I-Nh\u00f3m m\u00e1u": 26,
+    "I-N\u01a1i l\u00e0m vi\u1ec7c": 12,
+    "I-N\u01a1i sinh": 20,
+    "I-N\u01a1i \u1edf hi\u1ec7n t\u1ea1i": 32,
+    "I-Qu\u00ea qu\u00e1n": 22,
+    "I-Qu\u1ed1c t\u1ecbnh": 24,
+    "I-S\u1ed1 c\u0103n c\u01b0\u1edbc c\u00f4ng d\u00e2n": 28,
+    "I-S\u1ed1 h\u1ed9 chi\u1ebfu": 30,
+    "I-S\u1ed1 \u0111i\u1ec7n tho\u1ea1i": 6,
+    "I-S\u1edf th\u00edch": 46,
+    "I-Tr\u01b0\u1eddng h\u1ecdc": 40,
+    "I-Tu\u1ed5i": 54,
+    "I-T\u00ecnh tr\u1ea1ng h\u00f4n nh\u00e2n": 34,
+    "I-T\u00ednh c\u00e1ch": 48,
+    "I-T\u00f4n gi\u00e1o": 18,
+    "I-\u0110i\u1ec3m m\u1ea1nh": 50,
+    "I-\u0110i\u1ec3m y\u1ebfu": 52,
+    "I-\u0110\u1ecba ch\u1ec9": 10
+  },
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "bert",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "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",
+  "torch_dtype": "float32",
+  "transformers_version": "4.50.3",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 119547
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cae150939742dff5368c7edc3346038d7e6d34dacc7b39c6bec6dc3979ecef85
+size 709250084

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 +1,56 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "[PAD]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "100": {
+      "content": "[UNK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "101": {
+      "content": "[CLS]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "102": {
+      "content": "[SEP]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "103": {
+      "content": "[MASK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "clean_up_tokenization_spaces": false,
+  "cls_token": "[CLS]",
+  "do_lower_case": false,
+  "extra_special_tokens": {},
+  "mask_token": "[MASK]",
+  "model_max_length": 512,
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "strip_accents": null,
+  "tokenize_chinese_chars": true,
+  "tokenizer_class": "BertTokenizer",
+  "unk_token": "[UNK]"
+}