model improved

config.json

@@ -1198,8 +1198,7 @@
     "1184": "SYM",
     "1185": "VERB",
     "1186": "VERB+VERB",
-    "1187": "X",
-    "1188": "X+X"
+    "1187": "X"
   },
   "initializer_range": 0.02,
   "intermediate_size": 3072,
@@ -2391,8 +2390,7 @@
     "SYM": 1184,
     "VERB": 1185,
     "VERB+VERB": 1186,
-    "X": 1187,
-    "X+X": 1188
+    "X": 1187
   },
   "layer_norm_eps": 1e-05,
   "mask_token_id": 4,
@@ -2759,6 +2757,10 @@
           "\u0e40\u0e21\u0e37\u0e48\u0e2d",
           "\u0e21\u0e34\u0e25\u0e25\u0e34\u0e04\u0e31\u0e19"
         ],
+        "\u0e41\u0e01\u0e48\u0e19\u0e01\u0e19\u0e32\u0e07\u0e19\u0e27\u0e25": [
+          "\u0e41\u0e01\u0e48",
+          "\u0e19\u0e01\u0e19\u0e32\u0e07\u0e19\u0e27\u0e25"
+        ],
         "\u0e41\u0e01\u0e48\u0e19\u0e1e\u0e27\u0e23\u0e23\u0e13": [
           "\u0e41\u0e01\u0e48",
           "\u0e19\u0e1e\u0e27\u0e23\u0e23\u0e13"
@@ -10415,10 +10417,6 @@
           "\u0e41\u0e21\u0e49",
           "\u0e19\u0e32\u0e22"
         ],
-        "\u0e44\u0e14\u0e49\u0e41\u0e01\u0e48\u0e19\u0e32\u0e22": [
-          "\u0e44\u0e14\u0e49\u0e41\u0e01\u0e48",
-          "\u0e19\u0e32\u0e22"
-        ],
         "\u0e44\u0e21\u0e48\u0e27\u0e48\u0e32\u0e19\u0e31\u0e01\u0e28\u0e36\u0e01\u0e29\u0e32": [
           "\u0e44\u0e21\u0e48\u0e27\u0e48\u0e32",
           "\u0e19\u0e31\u0e01\u0e28\u0e36\u0e01\u0e29\u0e32"
@@ -10779,10 +10777,6 @@
           "\u0e41\u0e15\u0e48\u0e27\u0e48\u0e32",
           "\u0e19\u0e38\u0e49\u0e22"
         ],
-        "\u0e44\u0e14\u0e49\u0e41\u0e01\u0e48\u0e19\u0e01\u0e19\u0e32\u0e07\u0e19\u0e27\u0e25": [
-          "\u0e44\u0e14\u0e49\u0e41\u0e01\u0e48",
-          "\u0e19\u0e01\u0e19\u0e32\u0e07\u0e19\u0e27\u0e25"
-        ],
         "\u0e44\u0e21\u0e48\u0e27\u0e48\u0e32\u0e22\u0e39\u0e40\u0e19\u0e2a\u0e42\u0e01": [
           "\u0e44\u0e21\u0e48\u0e27\u0e48\u0e32",
           "\u0e22\u0e39\u0e40\u0e19\u0e2a\u0e42\u0e01"
@@ -36922,10 +36916,6 @@
           "\u0e2a\u0e38\u0e1e\u0e23\u0e23\u0e13\u0e0a\u0e19\u0e30\u0e1a\u0e38\u0e23\u0e35",
           "\u0e19\u0e32\u0e22\u0e01"
         ],
-        "\u0e2a\u0e38\u0e23\u0e2a\u0e31\u0e08\u0e08\u0e30": [
-          "\u0e2a\u0e38\u0e23\u0e2a\u0e31\u0e08\u0e08",
-          "\u0e30"
-        ],
         "\u0e2a\u0e38\u0e23\u0e32\u0e29\u0e0e\u0e23\u0e4c\u0e18\u0e32\u0e19\u0e35\u0e2d\u0e31\u0e19\u0e14\u0e31\u0e1a": [
           "\u0e2a\u0e38\u0e23\u0e32\u0e29\u0e0e\u0e23\u0e4c\u0e18\u0e32\u0e19\u0e35",
           "\u0e2d\u0e31\u0e19\u0e14\u0e31\u0e1a"
@@ -42037,10 +42027,6 @@
           "\u0e40\u0e2e\u0e0b\u0e1a\u0e2d\u0e25\u0e40\u0e25\u0e32\u0e30\u0e2b\u0e4c",
           "\u0e19\u0e31\u0e1a"
         ],
-        "\u0e40\u0e40\u0e1a\u0e07\u0e04\u0e4c\u0e2d\u0e32\u0e2a\u0e32": [
-          "\u0e40\u0e40\u0e1a\u0e07\u0e04\u0e4c",
-          "\u0e2d\u0e32\u0e2a\u0e32"
-        ],
         "\u0e41\u0e01\u0e49\u0e27\u0e2a\u0e23\u0e23\u0e0a\u0e35\u0e49": [
           "\u0e41\u0e01\u0e49\u0e27\u0e2a\u0e23\u0e23",
           "\u0e0a\u0e35\u0e49"
@@ -57708,10 +57694,6 @@
           "\u0e1b\u0e01\u0e15\u0e34",
           "\u0e23\u0e1f\u0e17."
         ],
-        "\u0e1b\u0e23\u0e30\u0e08\u0e33\u0e17\u0e32\u0e07": [
-          "\u0e1b\u0e23\u0e30\u0e08",
-          "\u0e33\u0e17\u0e32\u0e07"
-        ],
         "\u0e1e\u0e31\u0e12\u0e19\u0e32\u0e01\u0e27\u0e32\u0e07\u0e42\u0e08\u0e27": [
           "\u0e1e\u0e31\u0e12\u0e19\u0e32",
           "\u0e01\u0e27\u0e32\u0e07\u0e42\u0e08\u0e27"
@@ -59888,10 +59870,6 @@
           "\u0e23\u0e31\u0e01\u0e29\u0e32",
           "\u0e2a\u0e31\u0e07\u0e40\u0e01\u0e15"
         ],
-        "\u0e23\u0e31\u0e1a\u0e1c\u0e34\u0e14\u0e0a\u0e2d\u0e1a": [
-          "\u0e23",
-          "\u0e31\u0e1a\u0e1c\u0e34\u0e14\u0e0a\u0e2d\u0e1a"
-        ],
         "\u0e23\u0e34\u0e29\u0e22\u0e32\u0e19\u0e33\u0e1e\u0e32": [
           "\u0e23\u0e34\u0e29\u0e22\u0e32",
           "\u0e19\u0e33\u0e1e\u0e32"
@@ -61229,17 +61207,13 @@
         "\u0e41\u0e04\u0e0a\u0e0a\u0e48\u0e27\u0e25": [
           "\u0e41\u0e04\u0e0a",
           "\u0e0a\u0e48\u0e27\u0e25"
-        ],
-        "\u0e42\u0e1b\u0e23": [
-          "\u0e42",
-          "\u0e1b\u0e23"
         ]
       }
     }
   },
   "tokenizer_class": "BertTokenizerFast",
   "torch_dtype": "float32",
-  "transformers_version": "4.17.0",
+  "transformers_version": "4.40.1",
   "type_vocab_size": 1,
   "unk_token_id": 3,
   "use_cache": true,

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d5fd5781da463a0b138ac1ccce1bc3494bd54e3da314292b9c92115ea8e3e2c1
-size 396489009
+oid sha256:ef3caffbd115ba20be5aa8c8fb50610e00225b64391206516cf0062c041d1e11
+size 396469030

special_tokens_map.json

@@ -1 +1,7 @@
-{"unk_token": "<unk>", "sep_token": "</s>", "pad_token": "<pad>", "cls_token": "<s>", "mask_token": "<mask>"}
+{
+  "cls_token": "<s>",
+  "mask_token": "<mask>",
+  "pad_token": "<pad>",
+  "sep_token": "</s>",
+  "unk_token": "<unk>"
+}

supar.model

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:352ac506dab7446fa7722024a0c672ec9446af8578e2d53dee8a3998944cb9ae
-size 441815205
+oid sha256:3c1f6c41b25da9faeda132d228792fb5e9471954ab440391a4283099c6e4dd66
+size 442015286

tokenizer_config.json

@@ -1 +1,57 @@
-{"do_lower_case": false, "unk_token": "<unk>", "sep_token": "</s>", "pad_token": "<pad>", "cls_token": "<s>", "mask_token": "<mask>", "tokenize_chinese_chars": true, "strip_accents": false, "model_max_length": 416, "do_basic_tokenize": true, "never_split": null, "tokenizer_class": "BertTokenizer"}
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<pad>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "4": {
+      "content": "<mask>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "clean_up_tokenization_spaces": true,
+  "cls_token": "<s>",
+  "do_basic_tokenize": true,
+  "do_lower_case": false,
+  "mask_token": "<mask>",
+  "model_max_length": 416,
+  "never_split": null,
+  "pad_token": "<pad>",
+  "sep_token": "</s>",
+  "strip_accents": false,
+  "tokenize_chinese_chars": true,
+  "tokenizer_class": "BertTokenizerFast",
+  "unk_token": "<unk>"
+}