notebook: add demos for CoNLL-like exporter and Flair dataset loader
Browse files- Export-To-CoNLL.ipynb +76 -0
- FlairDatasetTest.ipynb +152 -0
Export-To-CoNLL.ipynb
ADDED
@@ -0,0 +1,76 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"cells": [
|
3 |
+
{
|
4 |
+
"cell_type": "code",
|
5 |
+
"execution_count": 1,
|
6 |
+
"id": "7b378ac1-1035-4f5d-9c16-68c14178389b",
|
7 |
+
"metadata": {},
|
8 |
+
"outputs": [],
|
9 |
+
"source": [
|
10 |
+
"from pathlib import Path\n",
|
11 |
+
"\n",
|
12 |
+
"from ftfy import fix_encoding"
|
13 |
+
]
|
14 |
+
},
|
15 |
+
{
|
16 |
+
"cell_type": "code",
|
17 |
+
"execution_count": 2,
|
18 |
+
"id": "1a8017c7-f7a1-4b17-804b-c4d404511969",
|
19 |
+
"metadata": {},
|
20 |
+
"outputs": [],
|
21 |
+
"source": [
|
22 |
+
"# Download and unzip archive from: https://www.dfki.uni-kl.de/cybermapping/data/CO-Fun-1.0-anonymized.zip\n",
|
23 |
+
"base_path = Path(\"./prepared-data-and-code/NER/CRF/\")\n",
|
24 |
+
"\n",
|
25 |
+
"splits = {\n",
|
26 |
+
" \"train\": base_path / \"train_set.txt\",\n",
|
27 |
+
" \"dev\": base_path / \"dev_set.txt\",\n",
|
28 |
+
" \"test\": base_path / \"test_set.txt\",\n",
|
29 |
+
"}"
|
30 |
+
]
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"cell_type": "code",
|
34 |
+
"execution_count": 3,
|
35 |
+
"id": "ab1deaef-adf3-4c4e-8da2-0bcc833615b3",
|
36 |
+
"metadata": {},
|
37 |
+
"outputs": [],
|
38 |
+
"source": [
|
39 |
+
"for split_name, split_file in splits.items():\n",
|
40 |
+
" with open(split_file, \"rt\") as f_p:\n",
|
41 |
+
" buffer = f_p.readlines()\n",
|
42 |
+
" buffer = \"[\" + \"\".join(buffer) + \"]\"\n",
|
43 |
+
" data = eval(buffer)\n",
|
44 |
+
"\n",
|
45 |
+
" with open(f\"{split_name}.tsv\", \"wt\") as f_out:\n",
|
46 |
+
" for sentence in data:\n",
|
47 |
+
" for line in sentence:\n",
|
48 |
+
" token, pos, ner = line\n",
|
49 |
+
" token = fix_encoding(token)\n",
|
50 |
+
" f_out.write(f\"{token}\\t{pos}\\t{ner}\\n\")\n",
|
51 |
+
" f_out.write(\"\\n\")"
|
52 |
+
]
|
53 |
+
}
|
54 |
+
],
|
55 |
+
"metadata": {
|
56 |
+
"kernelspec": {
|
57 |
+
"display_name": "Python 3 (ipykernel)",
|
58 |
+
"language": "python",
|
59 |
+
"name": "python3"
|
60 |
+
},
|
61 |
+
"language_info": {
|
62 |
+
"codemirror_mode": {
|
63 |
+
"name": "ipython",
|
64 |
+
"version": 3
|
65 |
+
},
|
66 |
+
"file_extension": ".py",
|
67 |
+
"mimetype": "text/x-python",
|
68 |
+
"name": "python",
|
69 |
+
"nbconvert_exporter": "python",
|
70 |
+
"pygments_lexer": "ipython3",
|
71 |
+
"version": "3.11.6"
|
72 |
+
}
|
73 |
+
},
|
74 |
+
"nbformat": 4,
|
75 |
+
"nbformat_minor": 5
|
76 |
+
}
|
FlairDatasetTest.ipynb
ADDED
@@ -0,0 +1,152 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"cells": [
|
3 |
+
{
|
4 |
+
"cell_type": "code",
|
5 |
+
"execution_count": 1,
|
6 |
+
"id": "87270bc7-2da0-490e-be8a-8cb06f38507b",
|
7 |
+
"metadata": {},
|
8 |
+
"outputs": [],
|
9 |
+
"source": [
|
10 |
+
"from flair.datasets.sequence_labeling import ColumnCorpus\n",
|
11 |
+
"\n",
|
12 |
+
"from pathlib import Path"
|
13 |
+
]
|
14 |
+
},
|
15 |
+
{
|
16 |
+
"cell_type": "code",
|
17 |
+
"execution_count": 2,
|
18 |
+
"id": "08ebe948-4179-4a75-8c6b-0e2fcad293d7",
|
19 |
+
"metadata": {},
|
20 |
+
"outputs": [],
|
21 |
+
"source": [
|
22 |
+
"columns = {0: \"text\", 2: \"ner\"}"
|
23 |
+
]
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"cell_type": "code",
|
27 |
+
"execution_count": 3,
|
28 |
+
"id": "af10c49b-002a-41f4-afde-172ac07251c0",
|
29 |
+
"metadata": {},
|
30 |
+
"outputs": [
|
31 |
+
{
|
32 |
+
"name": "stdout",
|
33 |
+
"output_type": "stream",
|
34 |
+
"text": [
|
35 |
+
"2024-03-25 12:57:38,583 Reading data from .\n",
|
36 |
+
"2024-03-25 12:57:38,585 Train: train.tsv\n",
|
37 |
+
"2024-03-25 12:57:38,586 Dev: dev.tsv\n",
|
38 |
+
"2024-03-25 12:57:38,586 Test: test.tsv\n"
|
39 |
+
]
|
40 |
+
}
|
41 |
+
],
|
42 |
+
"source": [
|
43 |
+
"corpus = ColumnCorpus(\".\",\n",
|
44 |
+
" column_format=columns,\n",
|
45 |
+
" train_file=\"train.tsv\",\n",
|
46 |
+
" dev_file=\"dev.tsv\",\n",
|
47 |
+
" test_file=\"test.tsv\",\n",
|
48 |
+
" comment_symbol=None)"
|
49 |
+
]
|
50 |
+
},
|
51 |
+
{
|
52 |
+
"cell_type": "code",
|
53 |
+
"execution_count": 4,
|
54 |
+
"id": "efd08709-9308-4244-a730-0d51b738520c",
|
55 |
+
"metadata": {},
|
56 |
+
"outputs": [
|
57 |
+
{
|
58 |
+
"data": {
|
59 |
+
"text/plain": [
|
60 |
+
"'Corpus: 758 train + 94 dev + 96 test sentences'"
|
61 |
+
]
|
62 |
+
},
|
63 |
+
"execution_count": 4,
|
64 |
+
"metadata": {},
|
65 |
+
"output_type": "execute_result"
|
66 |
+
}
|
67 |
+
],
|
68 |
+
"source": [
|
69 |
+
"str(corpus)"
|
70 |
+
]
|
71 |
+
},
|
72 |
+
{
|
73 |
+
"cell_type": "code",
|
74 |
+
"execution_count": 5,
|
75 |
+
"id": "6d857c95-ae9f-4aab-803a-0a7e03ff632d",
|
76 |
+
"metadata": {},
|
77 |
+
"outputs": [],
|
78 |
+
"source": [
|
79 |
+
"base_path = Path(\"./prepared-data-and-code/\")\n",
|
80 |
+
"\n",
|
81 |
+
"indices_for_split = {\n",
|
82 |
+
" \"train\": base_path / \"train_indices.txt\",\n",
|
83 |
+
" \"dev\": base_path / \"valid_indices.txt\",\n",
|
84 |
+
" \"test\": base_path / \"test_indices.txt\",\n",
|
85 |
+
"}"
|
86 |
+
]
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"cell_type": "code",
|
90 |
+
"execution_count": 6,
|
91 |
+
"id": "2529666f-793f-4d8c-a7aa-9b7d83975310",
|
92 |
+
"metadata": {},
|
93 |
+
"outputs": [],
|
94 |
+
"source": [
|
95 |
+
"def get_number_of_total_indices(filename: str) -> int:\n",
|
96 |
+
" indices_counter = 0\n",
|
97 |
+
" with open(filename, \"rt\") as f_p:\n",
|
98 |
+
" for line in f_p:\n",
|
99 |
+
" line = line.strip()\n",
|
100 |
+
" if not line:\n",
|
101 |
+
" continue\n",
|
102 |
+
" indices_counter += 1\n",
|
103 |
+
" return indices_counter"
|
104 |
+
]
|
105 |
+
},
|
106 |
+
{
|
107 |
+
"cell_type": "code",
|
108 |
+
"execution_count": 7,
|
109 |
+
"id": "49a69247-ef66-445b-b9e8-adcaac282e14",
|
110 |
+
"metadata": {},
|
111 |
+
"outputs": [],
|
112 |
+
"source": [
|
113 |
+
"train_indices = get_number_of_total_indices(indices_for_split[\"train\"])\n",
|
114 |
+
"dev_indices = get_number_of_total_indices(indices_for_split[\"dev\"])\n",
|
115 |
+
"test_indices = get_number_of_total_indices(indices_for_split[\"test\"])"
|
116 |
+
]
|
117 |
+
},
|
118 |
+
{
|
119 |
+
"cell_type": "code",
|
120 |
+
"execution_count": 8,
|
121 |
+
"id": "fb22489b-07d2-4403-a30c-6d92f00d252e",
|
122 |
+
"metadata": {},
|
123 |
+
"outputs": [],
|
124 |
+
"source": [
|
125 |
+
"assert len(corpus.train) == train_indices\n",
|
126 |
+
"assert len(corpus.dev) == dev_indices\n",
|
127 |
+
"assert len(corpus.test) == test_indices"
|
128 |
+
]
|
129 |
+
}
|
130 |
+
],
|
131 |
+
"metadata": {
|
132 |
+
"kernelspec": {
|
133 |
+
"display_name": "Python 3 (ipykernel)",
|
134 |
+
"language": "python",
|
135 |
+
"name": "python3"
|
136 |
+
},
|
137 |
+
"language_info": {
|
138 |
+
"codemirror_mode": {
|
139 |
+
"name": "ipython",
|
140 |
+
"version": 3
|
141 |
+
},
|
142 |
+
"file_extension": ".py",
|
143 |
+
"mimetype": "text/x-python",
|
144 |
+
"name": "python",
|
145 |
+
"nbconvert_exporter": "python",
|
146 |
+
"pygments_lexer": "ipython3",
|
147 |
+
"version": "3.11.6"
|
148 |
+
}
|
149 |
+
},
|
150 |
+
"nbformat": 4,
|
151 |
+
"nbformat_minor": 5
|
152 |
+
}
|