elishowk commited on
Commit
1e90333
·
1 Parent(s): 8743741
Files changed (1) hide show
  1. README.md +97 -8
README.md CHANGED
@@ -1,12 +1,101 @@
1
  ---
2
- license: miT
3
- languages: en
4
- tags:
5
- - exbert
6
- bigbang: 54.6
 
 
 
 
 
 
 
 
7
  ---
8
- # README
9
 
10
- Number[psor]=+Sing22~~éê$$
11
- ææER
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
12
 
 
 
1
  ---
2
+
3
+ language: eo
4
+
5
+ thumbnail: https://huggingface.co/blog/assets/01_how-to-train/EsperBERTo-thumbnail-v2.png
6
+
7
+ widget:
8
+
9
+ - text: "Jen la komenco de bela <mask>."
10
+
11
+ - text: "Uno du <mask>"
12
+
13
+ - text: "Jen finiĝas bela <mask>."
14
+
15
  ---
 
16
 
17
+ # EsperBERTo: RoBERTa-like Language model trained on Esperanto
18
+
19
+ **Companion model to blog post https://huggingface.co/blog/how-to-train** 🔥
20
+
21
+ ## Training Details
22
+
23
+ - current checkpoint: 566000
24
+
25
+ - machine name: `galinette`
26
+
27
+ ![](https://huggingface.co/blog/assets/01_how-to-train/EsperBERTo-thumbnail-v2.png)
28
+
29
+ ## Example pipeline
30
+
31
+ ```python
32
+
33
+ from transformers import pipeline
34
+
35
+ fill_mask = pipeline(
36
+
37
+ "fill-mask",
38
+
39
+ model="julien-c/EsperBERTo-small",
40
+
41
+ tokenizer="julien-c/EsperBERTo-small"
42
+
43
+ )
44
+
45
+ fill_mask("Jen la komenco de bela <mask>.")
46
+
47
+ # This is the beginning of a beautiful <mask>.
48
+
49
+ # =>
50
+
51
+ # {
52
+
53
+ # 'score':0.06502299010753632
54
+
55
+ # 'sequence':'<s> Jen la komenco de bela vivo.</s>'
56
+
57
+ # 'token':1099
58
+
59
+ # }
60
+
61
+ # {
62
+
63
+ # 'score':0.0421181358397007
64
+
65
+ # 'sequence':'<s> Jen la komenco de bela vespero.</s>'
66
+
67
+ # 'token':5100
68
+
69
+ # }
70
+
71
+ # {
72
+
73
+ # 'score':0.024884626269340515
74
+
75
+ # 'sequence':'<s> Jen la komenco de bela laboro.</s>'
76
+
77
+ # 'token':1570
78
+
79
+ # }
80
+
81
+ # {
82
+
83
+ # 'score':0.02324388362467289
84
+
85
+ # 'sequence':'<s> Jen la komenco de bela tago.</s>'
86
+
87
+ # 'token':1688
88
+
89
+ # }
90
+
91
+ # {
92
+
93
+ # 'score':0.020378097891807556
94
+
95
+ # 'sequence':'<s> Jen la komenco de bela festo.</s>'
96
+
97
+ # 'token':4580
98
+
99
+ # }
100
 
101
+ ```