add missing char tokens to vocab (with embeddings close to [UNK])

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "wietsedv/bert-base-dutch-cased",
+  "_name_or_path": "bert-base-dutch-cased",
   "architectures": [
     "BertForMaskedLM"
   ],
@@ -16,6 +16,9 @@
   "num_attention_heads": 12,
   "num_hidden_layers": 12,
   "pad_token_id": 3,
+  "position_embedding_type": "absolute",
+  "transformers_version": "4.5.1",
   "type_vocab_size": 2,
-  "vocab_size": 30000
-}
+  "use_cache": true,
+  "vocab_size": 30073
+}

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d6b583cdd24b3628d62041f927b783f432ba9f2ba1f203e7f02218e56ea22025
-size 436538834
+oid sha256:5ffe408c7eea0ffee4c257c6028f8c98146967e3ac3db51dba8e2bc8a4abddf5
+size 436761702

tf_model.h5

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:315fa75c371599484d77dedbadb0d49048d1aad9d8b2c56598403caa3b3f58c2
-size 532840208
+oid sha256:88cc47b929d21ed816d6ad8d5abea5c06ccae04a5f04f2d6b07da7d212aa18e1
+size 530923844

vocab.txt

@@ -29998,3 +29998,76 @@ zóó
 ##öl
 ##ön
 ##ör
+##Q
+##X
+##Ç
+##Ó
+##Ô
+##Ú
+##Û
+##Ü
+##à
+##á
+##â
+##ä
+##ê
+##ì
+##í
+##î
+##ñ
+##ò
+##ô
+##ù
+##ú
+##û
+##ü
+Q
+X
+a
+c
+e
+f
+g
+h
+i
+j
+k
+l
+m
+n
+o
+p
+q
+r
+s
+t
+u
+x
+y
+Ç
+Ó
+Ô
+Ú
+Û
+Ü
+à
+á
+â
+ä
+è
+é
+ê
+ë
+ì
+í
+î
+ï
+ñ
+ò
+ó
+ô
+ö
+ù
+ú
+û
+ü