model improved

config.json

@@ -1,4 +1,5 @@
 {
+  "_name_or_path": "KoichiYasuoka/roberta-small-japanese-aozora",
   "architectures": [
     "RobertaForTokenClassification"
   ],
@@ -10,108 +11,108 @@
   "hidden_dropout_prob": 0.1,
   "hidden_size": 256,
   "id2label": {
-    "0": "I-ADJ",
-    "1": "I-NUM",
-    "2": "I-ADV",
-    "3": "B-ADP",
-    "4": "I-NOUN",
-    "5": "I-DET",
-    "6": "B-ADV",
-    "7": "I-VERB",
-    "8": "VERB",
-    "9": "I-CCONJ",
-    "10": "AUX",
-    "11": "I-SCONJ",
-    "12": "B-DET",
-    "13": "B-ADJ",
-    "14": "I-PART",
-    "15": "ADV",
-    "16": "B-INTJ",
-    "17": "X",
-    "18": "CCONJ",
-    "19": "NOUN",
-    "20": "B-AUX",
-    "21": "INTJ",
-    "22": "B-SYM",
-    "23": "SCONJ",
-    "24": "I-INTJ",
-    "25": "ADJ",
-    "26": "B-PRON",
-    "27": "B-X",
-    "28": "DET",
-    "29": "SYM",
-    "30": "I-PRON",
-    "31": "I-SYM",
-    "32": "PUNCT",
-    "33": "I-PROPN",
-    "34": "I-AUX",
-    "35": "PART",
-    "36": "B-NUM",
-    "37": "I-X",
-    "38": "B-SCONJ",
-    "39": "B-PROPN",
-    "40": "B-PART",
-    "41": "I-ADP",
-    "42": "B-VERB",
-    "43": "ADP",
-    "44": "PROPN",
-    "45": "B-NOUN",
-    "46": "PRON",
-    "47": "B-CCONJ",
-    "48": "NUM"
+    "0": "SCONJ",
+    "1": "B-VERB",
+    "2": "I-SCONJ",
+    "3": "VERB",
+    "4": "B-AUX",
+    "5": "PART",
+    "6": "I-ADJ",
+    "7": "I-ADP",
+    "8": "B-NOUN",
+    "9": "PRON",
+    "10": "B-ADP",
+    "11": "NOUN",
+    "12": "NUM",
+    "13": "PUNCT",
+    "14": "B-X",
+    "15": "CCONJ",
+    "16": "ADV",
+    "17": "B-PRON",
+    "18": "I-PRON",
+    "19": "I-NOUN",
+    "20": "B-ADV",
+    "21": "AUX",
+    "22": "ADP",
+    "23": "I-NUM",
+    "24": "B-DET",
+    "25": "I-X",
+    "26": "B-PART",
+    "27": "I-DET",
+    "28": "B-INTJ",
+    "29": "INTJ",
+    "30": "B-NUM",
+    "31": "I-VERB",
+    "32": "I-ADV",
+    "33": "I-AUX",
+    "34": "B-PROPN",
+    "35": "B-SYM",
+    "36": "I-PART",
+    "37": "I-INTJ",
+    "38": "X",
+    "39": "I-PROPN",
+    "40": "B-ADJ",
+    "41": "DET",
+    "42": "SYM",
+    "43": "ADJ",
+    "44": "B-SCONJ",
+    "45": "B-CCONJ",
+    "46": "PROPN",
+    "47": "I-CCONJ",
+    "48": "I-SYM"
   },
   "initializer_range": 0.02,
   "intermediate_size": 768,
   "label2id": {
-    "ADJ": 25,
-    "ADP": 43,
-    "ADV": 15,
-    "AUX": 10,
-    "B-ADJ": 13,
-    "B-ADP": 3,
-    "B-ADV": 6,
-    "B-AUX": 20,
-    "B-CCONJ": 47,
-    "B-DET": 12,
-    "B-INTJ": 16,
-    "B-NOUN": 45,
-    "B-NUM": 36,
-    "B-PART": 40,
-    "B-PRON": 26,
-    "B-PROPN": 39,
-    "B-SCONJ": 38,
-    "B-SYM": 22,
-    "B-VERB": 42,
-    "B-X": 27,
-    "CCONJ": 18,
-    "DET": 28,
-    "I-ADJ": 0,
-    "I-ADP": 41,
-    "I-ADV": 2,
-    "I-AUX": 34,
-    "I-CCONJ": 9,
-    "I-DET": 5,
-    "I-INTJ": 24,
-    "I-NOUN": 4,
-    "I-NUM": 1,
-    "I-PART": 14,
-    "I-PRON": 30,
-    "I-PROPN": 33,
-    "I-SCONJ": 11,
-    "I-SYM": 31,
-    "I-VERB": 7,
-    "I-X": 37,
-    "INTJ": 21,
-    "NOUN": 19,
-    "NUM": 48,
-    "PART": 35,
-    "PRON": 46,
-    "PROPN": 44,
-    "PUNCT": 32,
-    "SCONJ": 23,
-    "SYM": 29,
-    "VERB": 8,
-    "X": 17
+    "ADJ": 43,
+    "ADP": 22,
+    "ADV": 16,
+    "AUX": 21,
+    "B-ADJ": 40,
+    "B-ADP": 10,
+    "B-ADV": 20,
+    "B-AUX": 4,
+    "B-CCONJ": 45,
+    "B-DET": 24,
+    "B-INTJ": 28,
+    "B-NOUN": 8,
+    "B-NUM": 30,
+    "B-PART": 26,
+    "B-PRON": 17,
+    "B-PROPN": 34,
+    "B-SCONJ": 44,
+    "B-SYM": 35,
+    "B-VERB": 1,
+    "B-X": 14,
+    "CCONJ": 15,
+    "DET": 41,
+    "I-ADJ": 6,
+    "I-ADP": 7,
+    "I-ADV": 32,
+    "I-AUX": 33,
+    "I-CCONJ": 47,
+    "I-DET": 27,
+    "I-INTJ": 37,
+    "I-NOUN": 19,
+    "I-NUM": 23,
+    "I-PART": 36,
+    "I-PRON": 18,
+    "I-PROPN": 39,
+    "I-SCONJ": 2,
+    "I-SYM": 48,
+    "I-VERB": 31,
+    "I-X": 25,
+    "INTJ": 29,
+    "NOUN": 11,
+    "NUM": 12,
+    "PART": 5,
+    "PRON": 9,
+    "PROPN": 46,
+    "PUNCT": 13,
+    "SCONJ": 0,
+    "SYM": 42,
+    "VERB": 3,
+    "X": 38
   },
   "layer_norm_eps": 1e-12,
   "max_position_embeddings": 128,
@@ -122,7 +123,7 @@
   "position_embedding_type": "absolute",
   "tokenizer_class": "RemBertTokenizerFast",
   "torch_dtype": "float32",
-  "transformers_version": "4.12.2",
+  "transformers_version": "4.11.3",
   "type_vocab_size": 2,
   "use_cache": true,
   "vocab_size": 250315