danameyer commited on
Commit
df9a8c8
·
verified ·
1 Parent(s): e5589f4

Add inference results

Browse files
README.md ADDED
@@ -0,0 +1,144 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ dataset_info:
3
+ config_name: default
4
+ features:
5
+ - name: project_name
6
+ dtype: large_string
7
+ - name: filename
8
+ dtype: large_string
9
+ - name: region_id
10
+ dtype: large_string
11
+ - name: line_id
12
+ dtype: large_string
13
+ - name: line_augmentation
14
+ dtype: large_string
15
+ - name: image
16
+ dtype:
17
+ struct:
18
+ - name: bytes
19
+ dtype: binary
20
+ - name: path
21
+ dtype: 'null'
22
+ - name: text
23
+ dtype: large_string
24
+ - name: line_reading_order
25
+ dtype: int64
26
+ - name: line_coords
27
+ dtype:
28
+ sequence:
29
+ sequence: int64
30
+ - name: line_baseline
31
+ dtype:
32
+ sequence:
33
+ sequence: int64
34
+ - name: region_reading_order
35
+ dtype: int64
36
+ - name: region_type
37
+ dtype: large_string
38
+ - name: region_coords
39
+ dtype:
40
+ sequence:
41
+ sequence: int64
42
+ - name: inference_20260605_220242_149959_model_dh-unibe_trocr-kurrent-XVI-XVII
43
+ dtype: string
44
+ - name: inference_20260608_164545_549548_model_dh-unibe_trocr-kurrent-XVI-XVII
45
+ dtype: string
46
+ splits:
47
+ - name: train
48
+ num_examples: 164
49
+ num_bytes: 34102936
50
+ download_size: 34102936
51
+ dataset_size: 34102936
52
+ configs:
53
+ - config_name: default
54
+ data_files:
55
+ - split: train
56
+ path: data/train/**/*.parquet
57
+ tags:
58
+ - image-to-text
59
+ - htr
60
+ - trocr
61
+ - inference
62
+ - pagexml
63
+ license: mit
64
+ ---
65
+
66
+ # Dataset Card for inference-test-dataset-workshop
67
+
68
+ This dataset is derived from [dh-unibe/kurrent-hanse-xvi-test-lines-with-inference](https://huggingface.co/datasets/dh-unibe/kurrent-hanse-xvi-test-lines-with-inference) and has been enriched with inference results.
69
+
70
+ ## Dataset Summary
71
+
72
+ This dataset contains 164 samples across 1 split(s).
73
+
74
+ ### Projects Included
75
+
76
+ 1505-02-10_Hanserezess,_Lübeck_Dienstag_nach_Scholastice_1505_(SAHST_Rep__2,_I_040-4)
77
+
78
+ ## Duplicate Line Information
79
+
80
+ Duplicate line statistics are calculated from the dataset key columns `filename`, `region_id`, and `line_id`, plus `project_name` when project metadata is available.
81
+
82
+ Only original rows are counted here.
83
+
84
+ - Duplicate rows: 0
85
+ - Duplicate groups: 0
86
+ - Duplicate excess rows: 0
87
+
88
+ ### Duplicate Lines by Split
89
+
90
+ - **train**: 0 duplicate rows, 0 duplicate groups, 0 duplicate excess rows
91
+
92
+ ## Dataset Structure
93
+
94
+ ### Data Splits
95
+
96
+ - **train**: 164 samples
97
+
98
+ ### Dataset Size
99
+
100
+ - Approximate total size: 32.52 MB
101
+ - Total samples: 164
102
+
103
+ ### Features
104
+
105
+ - **project_name**: `Value('large_string')`
106
+ - **filename**: `Value('large_string')`
107
+ - **region_id**: `Value('large_string')`
108
+ - **line_id**: `Value('large_string')`
109
+ - **line_augmentation**: `Value('large_string')`
110
+ - **image**: `{'bytes': Value('binary'), 'path': Value('null')}`
111
+ - **text**: `Value('large_string')`
112
+ - **line_reading_order**: `Value('int64')`
113
+ - **line_coords**: `List(List(Value('int64')))`
114
+ - **line_baseline**: `List(List(Value('int64')))`
115
+ - **region_reading_order**: `Value('int64')`
116
+ - **region_type**: `Value('large_string')`
117
+ - **region_coords**: `List(List(Value('int64')))`
118
+ - **inference_20260605_220242_149959_model_dh-unibe_trocr-kurrent-XVI-XVII**: `Value('string')`
119
+ - **inference_20260608_164545_549548_model_dh-unibe_trocr-kurrent-XVI-XVII**: `Value('string')`
120
+
121
+ ## Data Organization
122
+
123
+ Data is organized as parquet shards by split and project:
124
+
125
+ ```
126
+ data/
127
+ ├── <split>/
128
+ │ └── <project_name>/
129
+ │ └── <timestamp>-<shard>.parquet
130
+ ```
131
+
132
+ The HuggingFace Hub automatically merges all parquet files when loading the dataset.
133
+
134
+ ## Usage
135
+
136
+ ```python
137
+ from datasets import load_dataset
138
+
139
+ # Load entire dataset
140
+ dataset = load_dataset("danameyer/inference-test-dataset-workshop")
141
+
142
+ # Load specific split
143
+ dataset_split = load_dataset("danameyer/inference-test-dataset-workshop", split="train")
144
+ ```
data/train/1505-02-10_Hanserezess,_Lübeck_Dienstag_nach_Scholastice_1505_(SAHST_Rep__2,_I_040-4)/20260603_215209-0000.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aef7eef010dd0b27cbd2a84b903db09930ef7846eaff03d4d71aa36bb4c077f3
3
+ size 34102936