julien-c HF staff commited on
Commit
155c1d6
β€’
1 Parent(s): f9ad872

Migrate model card from transformers-repo

Browse files

Read announcement at https://discuss.huggingface.co/t/announcement-all-model-cards-will-be-migrated-to-hf-co-model-repos/2755
Original file history: https://github.com/huggingface/transformers/commits/master/model_cards/sachaarbonel/bert-italian-cased-finetuned-pos/README.md

Files changed (1) hide show
  1. README.md +96 -0
README.md ADDED
@@ -0,0 +1,96 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language: it
3
+ datasets:
4
+ - xtreme
5
+ ---
6
+
7
+ # Italian-Bert (Italian Bert) + POS πŸŽƒπŸ·
8
+
9
+ This model is a fine-tuned on [xtreme udpos Italian](https://huggingface.co/nlp/viewer/?dataset=xtreme&config=udpos.Italian) version of [Bert Base Italian](https://huggingface.co/dbmdz/bert-base-italian-cased) for **POS** downstream task.
10
+
11
+ ## Details of the downstream task (POS) - Dataset
12
+
13
+ - [Dataset: xtreme udpos Italian](https://huggingface.co/nlp/viewer/?dataset=xtreme&config=udpos.Italian) πŸ“š
14
+
15
+ | Dataset | # Examples |
16
+ | ---------------------- | ----- |
17
+ | Train | 716 K |
18
+ | Dev | 85 K |
19
+
20
+ - [Fine-tune on NER script provided by @stefan-it](https://raw.githubusercontent.com/stefan-it/fine-tuned-berts-seq/master/scripts/preprocess.py)
21
+
22
+ - Labels covered:
23
+
24
+ ```
25
+ ADJ
26
+ ADP
27
+ ADV
28
+ AUX
29
+ CCONJ
30
+ DET
31
+ INTJ
32
+ NOUN
33
+ NUM
34
+ PART
35
+ PRON
36
+ PROPN
37
+ PUNCT
38
+ SCONJ
39
+ SYM
40
+ VERB
41
+ X
42
+ ```
43
+
44
+ ## Metrics on evaluation set 🧾
45
+
46
+ | Metric | # score |
47
+ | :------------------------------------------------------------------------------------: | :-------: |
48
+ | F1 | **97.25**
49
+ | Precision | **97.15** |
50
+ | Recall | **97.36** |
51
+
52
+ ## Model in action πŸ”¨
53
+
54
+
55
+ Example of usage
56
+
57
+ ```python
58
+ from transformers import pipeline
59
+
60
+ nlp_pos = pipeline(
61
+ "ner",
62
+ model="sachaarbonel/bert-italian-cased-finetuned-pos",
63
+ tokenizer=(
64
+ 'sachaarbonel/bert-spanish-cased-finetuned-pos',
65
+ {"use_fast": False}
66
+ ))
67
+
68
+
69
+ text = 'Roma Γ¨ la Capitale d'Italia.'
70
+
71
+ nlp_pos(text)
72
+
73
+ '''
74
+ Output:
75
+ --------
76
+ [{'entity': 'PROPN', 'index': 1, 'score': 0.9995346665382385, 'word': 'roma'},
77
+ {'entity': 'AUX', 'index': 2, 'score': 0.9966597557067871, 'word': 'e'},
78
+ {'entity': 'DET', 'index': 3, 'score': 0.9994786977767944, 'word': 'la'},
79
+ {'entity': 'NOUN',
80
+ 'index': 4,
81
+ 'score': 0.9995198249816895,
82
+ 'word': 'capitale'},
83
+ {'entity': 'ADP', 'index': 5, 'score': 0.9990894198417664, 'word': 'd'},
84
+ {'entity': 'PART', 'index': 6, 'score': 0.57159024477005, 'word': "'"},
85
+ {'entity': 'PROPN',
86
+ 'index': 7,
87
+ 'score': 0.9994804263114929,
88
+ 'word': 'italia'},
89
+ {'entity': 'PUNCT', 'index': 8, 'score': 0.9772886633872986, 'word': '.'}]
90
+ '''
91
+ ```
92
+ Yeah! Not too bad πŸŽ‰
93
+
94
+ > Created by [Sacha Arbonel/@sachaarbonel](https://twitter.com/sachaarbonel) | [LinkedIn](https://www.linkedin.com/in/sacha-arbonel)
95
+
96
+ > Made with <span style="color: #e25555;">&hearts;</span> in Paris