model improved

config.json

@@ -14,102 +14,96 @@
     "2": "ADP",
     "3": "ADV",
     "4": "AUX",
-    "5": "AUX|Polarity=Neg",
-    "6": "AUX|Polarity=Neg+VERB",
-    "7": "B-ADJ",
-    "8": "B-ADJ+ADJ",
-    "9": "B-ADJ+AUX+AUX",
-    "10": "B-ADJ+VERB",
-    "11": "B-ADP",
-    "12": "B-ADP+NOUN",
-    "13": "B-ADV",
-    "14": "B-ADV+ADP",
-    "15": "B-ADV+ADP+VERB",
-    "16": "B-AUX",
-    "17": "B-AUX+AUX",
-    "18": "B-AUX+AUX+AUX",
-    "19": "B-AUX+AUX+PART",
-    "20": "B-AUX+SCONJ",
-    "21": "B-AUX|Polarity=Neg",
-    "22": "B-AUX|Polarity=Neg+VERB",
-    "23": "B-CCONJ",
-    "24": "B-DET",
-    "25": "B-INTJ",
-    "26": "B-NOUN",
-    "27": "B-NOUN+ADJ",
-    "28": "B-NOUN+ADP",
-    "29": "B-NOUN|Polarity=Neg",
-    "30": "B-NUM",
-    "31": "B-PART",
-    "32": "B-PRON",
-    "33": "B-PROPN",
-    "34": "B-PUNCT",
-    "35": "B-SCONJ",
-    "36": "B-SYM",
-    "37": "B-VERB",
-    "38": "B-VERB+AUX",
-    "39": "B-VERB+AUX+AUX",
-    "40": "B-VERB+AUX+NOUN",
-    "41": "B-VERB+AUX+PART",
-    "42": "B-VERB+AUX+SCONJ",
-    "43": "B-VERB+AUX|Polarity=Neg",
-    "44": "B-X",
-    "45": "CCONJ",
-    "46": "DET",
-    "47": "I-ADJ",
-    "48": "I-ADJ+ADJ",
-    "49": "I-ADJ+AUX+AUX",
-    "50": "I-ADJ+VERB",
-    "51": "I-ADP",
-    "52": "I-ADP+NOUN",
-    "53": "I-ADV",
-    "54": "I-ADV+ADP",
-    "55": "I-ADV+ADP+VERB",
-    "56": "I-AUX",
-    "57": "I-AUX+AUX",
-    "58": "I-AUX+AUX+AUX",
-    "59": "I-AUX+AUX+PART",
-    "60": "I-AUX+SCONJ",
-    "61": "I-AUX|Polarity=Neg",
-    "62": "I-AUX|Polarity=Neg+VERB",
-    "63": "I-CCONJ",
-    "64": "I-DET",
-    "65": "I-INTJ",
-    "66": "I-NOUN",
-    "67": "I-NOUN+ADJ",
-    "68": "I-NOUN+ADP",
-    "69": "I-NOUN|Polarity=Neg",
-    "70": "I-NUM",
-    "71": "I-PART",
-    "72": "I-PRON",
-    "73": "I-PROPN",
-    "74": "I-PUNCT",
-    "75": "I-SCONJ",
-    "76": "I-SYM",
-    "77": "I-VERB",
-    "78": "I-VERB+AUX",
-    "79": "I-VERB+AUX+AUX",
-    "80": "I-VERB+AUX+NOUN",
-    "81": "I-VERB+AUX+PART",
-    "82": "I-VERB+AUX+SCONJ",
-    "83": "I-VERB+AUX|Polarity=Neg",
-    "84": "I-X",
-    "85": "NOUN",
-    "86": "NOUN+ADP",
-    "87": "NOUN+AUX",
-    "88": "NUM",
-    "89": "PART",
-    "90": "PRON",
-    "91": "PRON+ADP",
-    "92": "PROPN",
-    "93": "PUNCT",
-    "94": "SCONJ",
-    "95": "SYM",
-    "96": "VERB",
-    "97": "VERB+AUX",
-    "98": "VERB+AUX+PART",
-    "99": "VERB+AUX|Polarity=Neg",
-    "100": "X"
+    "5": "AUX+VERB",
+    "6": "B-ADJ",
+    "7": "B-ADJ+ADJ",
+    "8": "B-ADJ+AUX+AUX",
+    "9": "B-ADJ+VERB",
+    "10": "B-ADP",
+    "11": "B-ADP+NOUN",
+    "12": "B-ADV",
+    "13": "B-ADV+ADP",
+    "14": "B-ADV+ADP+VERB",
+    "15": "B-AUX",
+    "16": "B-AUX+AUX",
+    "17": "B-AUX+AUX+AUX",
+    "18": "B-AUX+AUX+PART",
+    "19": "B-AUX+SCONJ",
+    "20": "B-AUX+VERB",
+    "21": "B-CCONJ",
+    "22": "B-DET",
+    "23": "B-INTJ",
+    "24": "B-NOUN",
+    "25": "B-NOUN+ADJ",
+    "26": "B-NOUN+ADP",
+    "27": "B-NUM",
+    "28": "B-PART",
+    "29": "B-PRON",
+    "30": "B-PROPN",
+    "31": "B-PUNCT+NOUN",
+    "32": "B-SCONJ",
+    "33": "B-SYM",
+    "34": "B-VERB",
+    "35": "B-VERB+AUX",
+    "36": "B-VERB+AUX+AUX",
+    "37": "B-VERB+AUX+NOUN",
+    "38": "B-VERB+AUX+PART",
+    "39": "B-VERB+AUX+SCONJ",
+    "40": "B-X",
+    "41": "CCONJ",
+    "42": "DET",
+    "43": "I-ADJ",
+    "44": "I-ADJ+ADJ",
+    "45": "I-ADJ+AUX+AUX",
+    "46": "I-ADJ+VERB",
+    "47": "I-ADP",
+    "48": "I-ADP+NOUN",
+    "49": "I-ADV",
+    "50": "I-ADV+ADP",
+    "51": "I-ADV+ADP+VERB",
+    "52": "I-AUX",
+    "53": "I-AUX+AUX",
+    "54": "I-AUX+AUX+AUX",
+    "55": "I-AUX+AUX+PART",
+    "56": "I-AUX+SCONJ",
+    "57": "I-AUX+VERB",
+    "58": "I-CCONJ",
+    "59": "I-DET",
+    "60": "I-INTJ",
+    "61": "I-NOUN",
+    "62": "I-NOUN+ADJ",
+    "63": "I-NOUN+ADP",
+    "64": "I-NUM",
+    "65": "I-PART",
+    "66": "I-PRON",
+    "67": "I-PROPN",
+    "68": "I-PUNCT+NOUN",
+    "69": "I-SCONJ",
+    "70": "I-SYM",
+    "71": "I-VERB",
+    "72": "I-VERB+AUX",
+    "73": "I-VERB+AUX+AUX",
+    "74": "I-VERB+AUX+NOUN",
+    "75": "I-VERB+AUX+PART",
+    "76": "I-VERB+AUX+SCONJ",
+    "77": "I-X",
+    "78": "NOUN",
+    "79": "NOUN+ADP",
+    "80": "NOUN+AUX",
+    "81": "NUM",
+    "82": "PART",
+    "83": "PRON",
+    "84": "PRON+ADP",
+    "85": "PROPN",
+    "86": "PUNCT",
+    "87": "PUNCT+PUNCT",
+    "88": "PUNCT+PUNCT+PUNCT",
+    "89": "SCONJ",
+    "90": "SYM",
+    "91": "VERB",
+    "92": "VERB+AUX",
+    "93": "VERB+AUX+PART",
+    "94": "X"
   },
   "initializer_range": 0.02,
   "intermediate_size": 3072,
@@ -119,102 +113,96 @@
     "ADP": 2,
     "ADV": 3,
     "AUX": 4,
-    "AUX|Polarity=Neg": 5,
-    "AUX|Polarity=Neg+VERB": 6,
-    "B-ADJ": 7,
-    "B-ADJ+ADJ": 8,
-    "B-ADJ+AUX+AUX": 9,
-    "B-ADJ+VERB": 10,
-    "B-ADP": 11,
-    "B-ADP+NOUN": 12,
-    "B-ADV": 13,
-    "B-ADV+ADP": 14,
-    "B-ADV+ADP+VERB": 15,
-    "B-AUX": 16,
-    "B-AUX+AUX": 17,
-    "B-AUX+AUX+AUX": 18,
-    "B-AUX+AUX+PART": 19,
-    "B-AUX+SCONJ": 20,
-    "B-AUX|Polarity=Neg": 21,
-    "B-AUX|Polarity=Neg+VERB": 22,
-    "B-CCONJ": 23,
-    "B-DET": 24,
-    "B-INTJ": 25,
-    "B-NOUN": 26,
-    "B-NOUN+ADJ": 27,
-    "B-NOUN+ADP": 28,
-    "B-NOUN|Polarity=Neg": 29,
-    "B-NUM": 30,
-    "B-PART": 31,
-    "B-PRON": 32,
-    "B-PROPN": 33,
-    "B-PUNCT": 34,
-    "B-SCONJ": 35,
-    "B-SYM": 36,
-    "B-VERB": 37,
-    "B-VERB+AUX": 38,
-    "B-VERB+AUX+AUX": 39,
-    "B-VERB+AUX+NOUN": 40,
-    "B-VERB+AUX+PART": 41,
-    "B-VERB+AUX+SCONJ": 42,
-    "B-VERB+AUX|Polarity=Neg": 43,
-    "B-X": 44,
-    "CCONJ": 45,
-    "DET": 46,
-    "I-ADJ": 47,
-    "I-ADJ+ADJ": 48,
-    "I-ADJ+AUX+AUX": 49,
-    "I-ADJ+VERB": 50,
-    "I-ADP": 51,
-    "I-ADP+NOUN": 52,
-    "I-ADV": 53,
-    "I-ADV+ADP": 54,
-    "I-ADV+ADP+VERB": 55,
-    "I-AUX": 56,
-    "I-AUX+AUX": 57,
-    "I-AUX+AUX+AUX": 58,
-    "I-AUX+AUX+PART": 59,
-    "I-AUX+SCONJ": 60,
-    "I-AUX|Polarity=Neg": 61,
-    "I-AUX|Polarity=Neg+VERB": 62,
-    "I-CCONJ": 63,
-    "I-DET": 64,
-    "I-INTJ": 65,
-    "I-NOUN": 66,
-    "I-NOUN+ADJ": 67,
-    "I-NOUN+ADP": 68,
-    "I-NOUN|Polarity=Neg": 69,
-    "I-NUM": 70,
-    "I-PART": 71,
-    "I-PRON": 72,
-    "I-PROPN": 73,
-    "I-PUNCT": 74,
-    "I-SCONJ": 75,
-    "I-SYM": 76,
-    "I-VERB": 77,
-    "I-VERB+AUX": 78,
-    "I-VERB+AUX+AUX": 79,
-    "I-VERB+AUX+NOUN": 80,
-    "I-VERB+AUX+PART": 81,
-    "I-VERB+AUX+SCONJ": 82,
-    "I-VERB+AUX|Polarity=Neg": 83,
-    "I-X": 84,
-    "NOUN": 85,
-    "NOUN+ADP": 86,
-    "NOUN+AUX": 87,
-    "NUM": 88,
-    "PART": 89,
-    "PRON": 90,
-    "PRON+ADP": 91,
-    "PROPN": 92,
-    "PUNCT": 93,
-    "SCONJ": 94,
-    "SYM": 95,
-    "VERB": 96,
-    "VERB+AUX": 97,
-    "VERB+AUX+PART": 98,
-    "VERB+AUX|Polarity=Neg": 99,
-    "X": 100
+    "AUX+VERB": 5,
+    "B-ADJ": 6,
+    "B-ADJ+ADJ": 7,
+    "B-ADJ+AUX+AUX": 8,
+    "B-ADJ+VERB": 9,
+    "B-ADP": 10,
+    "B-ADP+NOUN": 11,
+    "B-ADV": 12,
+    "B-ADV+ADP": 13,
+    "B-ADV+ADP+VERB": 14,
+    "B-AUX": 15,
+    "B-AUX+AUX": 16,
+    "B-AUX+AUX+AUX": 17,
+    "B-AUX+AUX+PART": 18,
+    "B-AUX+SCONJ": 19,
+    "B-AUX+VERB": 20,
+    "B-CCONJ": 21,
+    "B-DET": 22,
+    "B-INTJ": 23,
+    "B-NOUN": 24,
+    "B-NOUN+ADJ": 25,
+    "B-NOUN+ADP": 26,
+    "B-NUM": 27,
+    "B-PART": 28,
+    "B-PRON": 29,
+    "B-PROPN": 30,
+    "B-PUNCT+NOUN": 31,
+    "B-SCONJ": 32,
+    "B-SYM": 33,
+    "B-VERB": 34,
+    "B-VERB+AUX": 35,
+    "B-VERB+AUX+AUX": 36,
+    "B-VERB+AUX+NOUN": 37,
+    "B-VERB+AUX+PART": 38,
+    "B-VERB+AUX+SCONJ": 39,
+    "B-X": 40,
+    "CCONJ": 41,
+    "DET": 42,
+    "I-ADJ": 43,
+    "I-ADJ+ADJ": 44,
+    "I-ADJ+AUX+AUX": 45,
+    "I-ADJ+VERB": 46,
+    "I-ADP": 47,
+    "I-ADP+NOUN": 48,
+    "I-ADV": 49,
+    "I-ADV+ADP": 50,
+    "I-ADV+ADP+VERB": 51,
+    "I-AUX": 52,
+    "I-AUX+AUX": 53,
+    "I-AUX+AUX+AUX": 54,
+    "I-AUX+AUX+PART": 55,
+    "I-AUX+SCONJ": 56,
+    "I-AUX+VERB": 57,
+    "I-CCONJ": 58,
+    "I-DET": 59,
+    "I-INTJ": 60,
+    "I-NOUN": 61,
+    "I-NOUN+ADJ": 62,
+    "I-NOUN+ADP": 63,
+    "I-NUM": 64,
+    "I-PART": 65,
+    "I-PRON": 66,
+    "I-PROPN": 67,
+    "I-PUNCT+NOUN": 68,
+    "I-SCONJ": 69,
+    "I-SYM": 70,
+    "I-VERB": 71,
+    "I-VERB+AUX": 72,
+    "I-VERB+AUX+AUX": 73,
+    "I-VERB+AUX+NOUN": 74,
+    "I-VERB+AUX+PART": 75,
+    "I-VERB+AUX+SCONJ": 76,
+    "I-X": 77,
+    "NOUN": 78,
+    "NOUN+ADP": 79,
+    "NOUN+AUX": 80,
+    "NUM": 81,
+    "PART": 82,
+    "PRON": 83,
+    "PRON+ADP": 84,
+    "PROPN": 85,
+    "PUNCT": 86,
+    "PUNCT+PUNCT": 87,
+    "PUNCT+PUNCT+PUNCT": 88,
+    "SCONJ": 89,
+    "SYM": 90,
+    "VERB": 91,
+    "VERB+AUX": 92,
+    "VERB+AUX+PART": 93,
+    "X": 94
   },
   "layer_norm_eps": 1e-07,
   "max_position_embeddings": 512,
@@ -226,9 +214,12 @@
   "pooler_dropout": 0,
   "pooler_hidden_act": "gelu",
   "pooler_hidden_size": 768,
-  "pos_att_type": null,
-  "position_biased_input": true,
-  "relative_attention": false,
+  "pos_att_type": [
+    "p2c",
+    "c2p"
+  ],
+  "position_biased_input": false,
+  "relative_attention": true,
   "task_specific_params": {
     "upos_multiword": {
       "ADJ+ADJ": {
@@ -318,7 +309,7 @@
           "\u3051\u3069"
         ]
       },
-      "AUX|Polarity=Neg+VERB": {
+      "AUX+VERB": {
         "\u306a\u304f\u306a\u3063": [
           "\u306a\u304f",
           "\u306a\u3063"
@@ -380,6 +371,18 @@
           "\u304b"
         ]
       },
+      "PUNCT+NOUN": {
+        ",\u201c\u62c9\u81f4\u76e3\u7981\u201d\u554f\u984c": [
+          ",",
+          "\u201c\u62c9\u81f4\u76e3\u7981\u201d\u554f\u984c"
+        ]
+      },
+      "PUNCT+PUNCT": {
+        ",\u201c": [
+          ",",
+          "\u201c"
+        ]
+      },
       "VERB+AUX": {
         "\u3044\u308f\u308c": [
           "\u3044\u308f",
@@ -393,6 +396,10 @@
           "\u304a\u3053\u306a\u308f",
           "\u308c"
         ],
+        "\u304b\u307e\u308f\u305a": [
+          "\u304b\u307e\u308f",
+          "\u305a"
+        ],
         "\u3055\u3089\u308f\u308c": [
           "\u3055\u3089\u308f",
           "\u308c"
@@ -417,6 +424,10 @@
           "\u306a\u3058\u307e",
           "\u305b"
         ],
+        "\u306a\u3089\u305a": [
+          "\u306a\u3089",
+          "\u305a"
+        ],
         "\u307f\u306a\u3055\u308c": [
           "\u307f\u306a\u3055",
           "\u308c"
@@ -469,6 +480,10 @@
           "\u548c\u307e",
           "\u305b"
         ],
+        "\u554f\u308f\u305a": [
+          "\u554f\u308f",
+          "\u305a"
+        ],
         "\u554f\u308f\u308c": [
           "\u554f\u308f",
           "\u308c"
@@ -537,6 +552,10 @@
           "\u679c\u305f\u3059",
           "\u3079\u304f"
         ],
+        "\u69cb\u308f\u305a": [
+          "\u69cb\u308f",
+          "\u305a"
+        ],
         "\u6b4c\u308f\u308c": [
           "\u6b4c\u308f",
           "\u308c"
@@ -577,6 +596,10 @@
           "\u884c\u3063",
           "\u3061\u3083\u3044"
         ],
+        "\u884c\u308f\u305a": [
+          "\u884c\u308f",
+          "\u305a"
+        ],
         "\u884c\u308f\u308c": [
           "\u884c\u308f",
           "\u308c"
@@ -719,34 +742,12 @@
           "\u305f",
           "\u305f\u3081\u306b"
         ]
-      },
-      "VERB+AUX|Polarity=Neg": {
-        "\u304b\u307e\u308f\u305a": [
-          "\u304b\u307e\u308f",
-          "\u305a"
-        ],
-        "\u306a\u3089\u305a": [
-          "\u306a\u3089",
-          "\u305a"
-        ],
-        "\u554f\u308f\u305a": [
-          "\u554f\u308f",
-          "\u305a"
-        ],
-        "\u69cb\u308f\u305a": [
-          "\u69cb\u308f",
-          "\u305a"
-        ],
-        "\u884c\u308f\u305a": [
-          "\u884c\u308f",
-          "\u305a"
-        ]
       }
     }
   },
   "tokenizer_class": "DebertaV2TokenizerFast",
   "torch_dtype": "float32",
-  "transformers_version": "4.19.4",
+  "transformers_version": "4.22.1",
   "type_vocab_size": 0,
   "vocab_size": 32000
 }

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:37128693dcce708756bf50fbee84d5d9d61bf02b3c20aa56044b88b1e512d862
-size 440482611
+oid sha256:449963d8f2f70cb7606c6f7a8ec258c9f2d0ace75eacf31d48e356136ceb9cee
+size 498750547

special_tokens_map.json

@@ -1 +1,9 @@
-{"bos_token": "[CLS]", "eos_token": "[SEP]", "unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]"}
+{
+  "bos_token": "[CLS]",
+  "cls_token": "[CLS]",
+  "eos_token": "[SEP]",
+  "mask_token": "[MASK]",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "unk_token": "[UNK]"
+}

supar.model

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ddd1931e9712901e6b03586f19cb7ba4ab7bb3da1f40af69b99d77cdc9d5da48
-size 488947819
+oid sha256:bf053426925c490ebf2a2a0bc6465db9726f7359f3425833e6a5691979032caf
+size 547274827

tokenizer_config.json

@@ -1 +1,14 @@
-{"do_lower_case": false, "bos_token": "[CLS]", "eos_token": "[SEP]", "unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]", "split_by_punct": true, "keep_accents": true, "model_max_length": 512, "tokenizer_class": "DebertaV2TokenizerFast"}
+{
+  "bos_token": "[CLS]",
+  "cls_token": "[CLS]",
+  "do_lower_case": false,
+  "eos_token": "[SEP]",
+  "keep_accents": true,
+  "mask_token": "[MASK]",
+  "model_max_length": 512,
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "split_by_punct": true,
+  "tokenizer_class": "DebertaV2TokenizerFast",
+  "unk_token": "[UNK]"
+}