readme: update model card

README.md

@@ -3,10 +3,10 @@ tags:
 - flair
 - token-classification
 - sequence-tagger-model
-language: 
-- en 
-- de 
-- fr 
+language:
+- en
+- de
+- fr
 - it
 - nl
 - pl
@@ -26,7 +26,7 @@ widget:
 
 This is the default multilingual universal part-of-speech tagging model that ships with [Flair](https://github.com/flairNLP/flair/).
 
-F1-Score: **98,47** (12 UD Treebanks covering English, German, French, Italian, Dutch, Polish, Spanish, Swedish, Danish, Norwegian, Finnish and Czech)
+F1-Score: **96.87** (12 UD Treebanks covering English, German, French, Italian, Dutch, Polish, Spanish, Swedish, Danish, Norwegian, Finnish and Czech)
 
 Predicts universal POS tags:
 
@@ -94,14 +94,14 @@ Token[6]: "say" → VERB (0.9998)
 Token[7]: "." → PUNCT (1.0)
 ```
 
-So, the words "*Ich*" and "*they*" are labeled as **pronouns** (PRON), while "*liebe*" and "*say*" are labeled as **verbs** (VERB) in the multilingual sentence "*Ich liebe Berlin, as they say*". 
+So, the words "*Ich*" and "*they*" are labeled as **pronouns** (PRON), while "*liebe*" and "*say*" are labeled as **verbs** (VERB) in the multilingual sentence "*Ich liebe Berlin, as they say*".
 
 
 ---
 
 ### Training: Script to train this model
 
-The following Flair script was used to train this model: 
+The following Flair script was used to train this model:
 
 ```python
 from flair.data import MultiCorpus
@@ -129,11 +129,10 @@ corpus = MultiCorpus([
 tag_type = 'upos'
 
 # 3. make the tag dictionary from the corpus
-tag_dictionary = corpus.make_tag_dictionary(tag_type=tag_type)
+tag_dictionary = corpus.make_label_dictionary(label_type=tag_type)
 
 # 4. initialize each embedding we use
 embedding_types = [
-
     # contextual string embeddings, forward
     FlairEmbeddings('multi-forward'),
 
@@ -141,7 +140,7 @@ embedding_types = [
     FlairEmbeddings('multi-backward'),
 ]
 
-# embedding stack consists of Flair and GloVe embeddings
+# embedding stack consists of Flair embeddings
 embeddings = StackedEmbeddings(embeddings=embedding_types)
 
 # 5. initialize sequence tagger