-
Notifications
You must be signed in to change notification settings - Fork 41
/
test_backend_fasttext.py
217 lines (177 loc) · 6.54 KB
/
test_backend_fasttext.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
"""Unit tests for the fastText backend in Annif"""
import logging
import pytest
import annif.backend
import annif.corpus
from annif.exception import NotSupportedException
fasttext = pytest.importorskip("annif.backend.fasttext")
def test_fasttext_default_params(project):
fasttext_type = annif.backend.get_backend("fasttext")
fasttext = fasttext_type(backend_id="fasttext", config_params={}, project=project)
expected_default_params = {
"limit": 100,
"chunksize": 1,
"dim": 100,
"lr": 0.25,
"epoch": 5,
"loss": "hs",
}
actual_params = fasttext.params
for param, val in expected_default_params.items():
assert param in actual_params and actual_params[param] == val
def test_fasttext_train(document_corpus, project, datadir):
fasttext_type = annif.backend.get_backend("fasttext")
fasttext = fasttext_type(
backend_id="fasttext",
config_params={"limit": 50, "dim": 100, "lr": 0.25, "epoch": 20, "loss": "hs"},
project=project,
)
fasttext.train(document_corpus)
assert fasttext._model is not None
assert datadir.join("fasttext-model").exists()
assert datadir.join("fasttext-model").size() > 0
def test_fasttext_train_cached_jobs(project, datadir):
assert datadir.join("fasttext-train.txt").exists()
datadir.join("fasttext-model").remove()
fasttext_type = annif.backend.get_backend("fasttext")
fasttext = fasttext_type(
backend_id="fasttext",
config_params={"limit": 50, "dim": 100, "lr": 0.25, "epoch": 20, "loss": "hs"},
project=project,
)
fasttext.train("cached", jobs=2)
assert fasttext._model is not None
assert datadir.join("fasttext-model").exists()
assert datadir.join("fasttext-model").size() > 0
def test_fasttext_train_unknown_subject(tmpdir, datadir, project):
fasttext_type = annif.backend.get_backend("fasttext")
fasttext = fasttext_type(
backend_id="fasttext",
config_params={"limit": 50, "dim": 100, "lr": 0.25, "epoch": 20, "loss": "hs"},
project=project,
)
tmpfile = tmpdir.join("document.tsv")
tmpfile.write(
"nonexistent\thttp://example.com/nonexistent\n"
+ "arkeologia\thttp://www.yso.fi/onto/yso/p1265"
)
document_corpus = annif.corpus.DocumentFile(str(tmpfile), project.subjects)
fasttext.train(document_corpus)
assert fasttext._model is not None
assert datadir.join("fasttext-model").exists()
assert datadir.join("fasttext-model").size() > 0
def test_fasttext_train_nodocuments(project, empty_corpus):
fasttext_type = annif.backend.get_backend("fasttext")
fasttext = fasttext_type(
backend_id="fasttext",
config_params={"limit": 50, "dim": 100, "lr": 0.25, "epoch": 20, "loss": "hs"},
project=project,
)
with pytest.raises(NotSupportedException) as excinfo:
fasttext.train(empty_corpus)
assert "training backend fasttext with no documents" in str(excinfo.value)
def test_train_fasttext_params(document_corpus, project, caplog):
logger = annif.logger
logger.propagate = True
fasttext_type = annif.backend.get_backend("fasttext")
fasttext = fasttext_type(
backend_id="fasttext",
config_params={"limit": 51, "dim": 101, "lr": 0.21, "epoch": 21, "loss": "hs"},
project=project,
)
params = {"dim": 1, "lr": 42.1, "epoch": 0}
with caplog.at_level(logging.DEBUG, logger="annif"):
fasttext.train(document_corpus, params)
parameters_heading = "Backend fasttext: Model parameters:"
assert parameters_heading in caplog.text
for line in caplog.text.splitlines():
if parameters_heading in line:
assert "'dim': 1" in line
assert "'lr': 42.1" in line
assert "'epoch': 0" in line
def test_fasttext_train_pretrained(
datadir, document_corpus, project, pretrained_vectors
):
assert pretrained_vectors.exists()
assert pretrained_vectors.size() > 0
fasttext_type = annif.backend.get_backend("fasttext")
fasttext = fasttext_type(
backend_id="fasttext",
config_params={
"limit": 50,
"dim": 100,
"lr": 0.25,
"epoch": 20,
"loss": "hs",
"pretrainedVectors": str(pretrained_vectors),
},
project=project,
)
fasttext.train(document_corpus)
assert fasttext._model is not None
assert datadir.join("fasttext-model").exists()
assert datadir.join("fasttext-model").size() > 0
def test_fasttext_train_pretrained_wrong_dim(
datadir, document_corpus, project, pretrained_vectors
):
assert pretrained_vectors.exists()
assert pretrained_vectors.size() > 0
fasttext_type = annif.backend.get_backend("fasttext")
fasttext = fasttext_type(
backend_id="fasttext",
config_params={
"limit": 50,
"dim": 50,
"lr": 0.25,
"epoch": 20,
"loss": "hs",
"pretrainedVectors": str(pretrained_vectors),
},
project=project,
)
with pytest.raises(ValueError):
fasttext.train(document_corpus)
assert fasttext._model is None
def test_fasttext_suggest(project):
fasttext_type = annif.backend.get_backend("fasttext")
fasttext = fasttext_type(
backend_id="fasttext",
config_params={
"limit": 50,
"chunksize": 1,
"dim": 100,
"lr": 0.25,
"epoch": 20,
"loss": "hs",
},
project=project,
)
results = fasttext.suggest(
[
"""Arkeologiaa sanotaan joskus myös
muinaistutkimukseksi tai muinaistieteeksi. Se on humanistinen tiede
tai oikeammin joukko tieteitä, jotka tutkivat ihmisen menneisyyttä.
Tutkimusta tehdään analysoimalla muinaisjäännöksiä eli niitä jälkiä,
joita ihmisten toiminta on jättänyt maaperään tai vesistöjen
pohjaan."""
]
)[0]
assert len(results) > 0
archaeology = project.subjects.by_uri("http://www.yso.fi/onto/yso/p1265")
assert archaeology in [result.subject_id for result in results]
def test_fasttext_suggest_empty_chunks(project):
fasttext_type = annif.backend.get_backend("fasttext")
fasttext = fasttext_type(
backend_id="fasttext",
config_params={
"limit": 50,
"chunksize": 1,
"dim": 100,
"lr": 0.25,
"epoch": 20,
"loss": "hs",
},
project=project,
)
results = fasttext.suggest([""])[0]
assert len(results) == 0