thodel commited on
Commit
ff805db
1 Parent(s): 65460ed

Update README.md

Browse files

edit readme for clarity abt. material used

Files changed (1) hide show
  1. README.md +7 -58
README.md CHANGED
@@ -9,6 +9,7 @@ tags:
9
  - kurrent
10
  - ocr
11
  - htr
 
12
  ---
13
  # TrOCR Kurrent-Model 19th century
14
  Base model: **microsoft/trocr-base-handwritten**
@@ -22,61 +23,9 @@ Eval CER: 0.02827
22
  Test CER: 0.02655
23
 
24
  Finetuned on Kurrent-dataset, containing:
25
- - Weber_Leo
26
- - Nekrolog
27
- - Nekrolog_2
28
- - Huber,_Eugen,Tagebuch_in_Briefform,_1910_Hand
29
- - Semper_reinschriften_2
30
- - HUMBOLDT_PATZIG_duplicated
31
- - Konzilsprotokolle_A_Haeckermann_2
32
- - Konzilsprotokolle_A_Haeckermann_3
33
- - Nekrolog_M2
34
- - Konzilsprotokolle_B_Schwartz_10_2017
35
- - HUMBOLDT_PATZIG
36
- - Todesurteile_1891
37
- - Barlach_M3
38
- - Nekrolog_M3
39
- - Todesurteile_1849+1891
40
- - Konzilsprotokolle_B_Schwartz
41
- - Pyl_M2
42
- - St_Aegidi_M1
43
- - Stammler_Rudolf
44
- - 232_duplicated
45
- - StAHZ_Training
46
- - Todesurteile_1849
47
- - DreierM2
48
- - Konzilsprotokolle_B_Schwartz_ext
49
- - Robert_Musil_M1
50
- - Pyl_M3
51
- - Todesurteile
52
- - Pyl_M1
53
- - GemeindeChronikM1
54
- - Lohde_Handschriften_digital
55
- - Binder_Kochbuch
56
- - Binder_Kochbuch_M2
57
- - Hendschel_M3
58
- - Testdaten_öbf_1
59
- - 231_duplicated
60
- - Pyl_M4
61
- - Huber,_Eugen,Tagebuch_in_Briefform,_1911_Hand
62
- - semper-eigeneHand_t2i-versuch5_duplicated
63
- - Protokoll_Hoftheater_1806_M1
64
- - Huber,_Eugen,Tagebuch_in_Briefform,_1912_Hand
65
- - Briefe_Eugen_Huber_an_Max_Rümelin
66
- - HHStA-KK_M2_duplicated
67
- - Polizeiberichte_M3_duplicated
68
- - Dreier_M1
69
- - Mulinen_Helene_von
70
- - Schweiz__Frauenverein
71
- - semper-schreiber_t2i_versuch6
72
- - Konzilsprotokolle_A_Haeckermann
73
- - tuempel_M1
74
- - Huber,_Eugen,Tagebuch_in_Briefform,_1913_T2I,_Hand
75
- - German_Kurrent_(Reichsgericht)_duplicated
76
- - Burckhardt_Walter
77
- - Urtheil_In_einer_Auseinandersetzung_über_Pensionszahlung_zwischen_dem_Kanzleirat_Althof_und_dem_Fürstlich_Lippischen_Ministerium_Ohne_Ort,_25_1_1856_-_1_2_1856
78
- - Huber,_Absender_Rümelin,_Max
79
- - Hendschel_ext_M1
80
- - Bridel_Louis
81
- - Pyl_M5
82
- - Semper_t2i-weitere
 
9
  - kurrent
10
  - ocr
11
  - htr
12
+ - 19th century
13
  ---
14
  # TrOCR Kurrent-Model 19th century
15
  Base model: **microsoft/trocr-base-handwritten**
 
23
  Test CER: 0.02655
24
 
25
  Finetuned on Kurrent-dataset, containing:
26
+ - Material from the State Archives of Zurich ("Regierungsratsprotokolle"), provided by the State Archives of Zurich
27
+ - Lecture notes of Humboldt Lectures, provided by the Berlin-Brandenburgian Academy of Sciences
28
+ - Diary of Eugen Huber, provided by the University of Zurich
29
+ - Handwritting and Copies by and of Gottfried Semper
30
+ - Konzilsprotokolle, University of Greifswald (19th century)
31
+ - as well as many other smaller collections/examples