Hugging Face's logo

mbley
/
german-webtext-quality-classifier-base

Text Classification
setfit
Safetensors
sentence-transformers
German
distilbert
generated_from_setfit_trainer
Model card Files
xet
Community
mbley commited on
Commit
c969ff2
·
verified ·
1 Parent(s): 80dbde5

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +2 -174
README.md CHANGED
@@ -30,11 +30,12 @@ base_model:
30
  # Bootstrapping a Sentence-Level Corpus Quality Classifier for Web Text using Active Learning (RANLP25)
31
 
32
  A multi-label sentence classifier trained with Active Learning for predicting high- or low-qality labels of german webtext.
 
33
  Training and evaluation code: <https://github.com/maximilian-bley/german-webtext-quality-classification>
34
 
35
  ## Model Details
36
 
37
- - **Labels**
38
 
39
  - **0=Sentence Boundary:** Sentence boundary errors occur if the start or ending of a sentence is malformed. This is the case if it begins with a lower case letter or an atypical character, or lacks a proper terminal punctuation mark (e.g., period, exclamation mark, or question mark).
40
 
@@ -111,7 +112,6 @@ preds = model("在 Greding 出 口 离 开 A9 高 速 公 路 。")
111
 
112
  ## Training Details
113
 
114
-
115
  ### Training Hyperparameters
116
  - batch_size: (8, 8)
117
  - num_epochs: (1, 16)
@@ -131,178 +131,6 @@ preds = model("在 Greding 出 口 离 开 A9 高 速 公 路 。")
131
  - eval_max_steps: -1
132
  - load_best_model_at_end: False
133
 
134
- ### Training Results
135
- | Epoch | Step | Training Loss | Validation Loss |
136
- |:------:|:-----:|:-------------:|:---------------:|
137
- | 0.0001 | 1 | 4.5018 | - |
138
- | 0.0060 | 100 | 5.2045 | - |
139
- | 0.0119 | 200 | 4.559 | - |
140
- | 0.0179 | 300 | 3.4579 | - |
141
- | 0.0239 | 400 | 3.106 | - |
142
- | 0.0298 | 500 | 2.7464 | - |
143
- | 0.0358 | 600 | 2.5813 | - |
144
- | 0.0417 | 700 | 2.5341 | - |
145
- | 0.0477 | 800 | 2.5279 | - |
146
- | 0.0537 | 900 | 2.361 | - |
147
- | 0.0596 | 1000 | 2.2318 | - |
148
- | 0.0656 | 1100 | 1.8437 | - |
149
- | 0.0716 | 1200 | 1.6423 | - |
150
- | 0.0775 | 1300 | 1.7572 | - |
151
- | 0.0835 | 1400 | 1.8163 | - |
152
- | 0.0895 | 1500 | 1.4293 | - |
153
- | 0.0954 | 1600 | 1.3842 | - |
154
- | 0.1014 | 1700 | 0.9845 | - |
155
- | 0.1073 | 1800 | 1.0666 | - |
156
- | 0.1133 | 1900 | 0.6876 | - |
157
- | 0.1193 | 2000 | 1.4398 | - |
158
- | 0.1252 | 2100 | 0.7268 | - |
159
- | 0.1312 | 2200 | 0.7272 | - |
160
- | 0.1372 | 2300 | 0.9801 | - |
161
- | 0.1431 | 2400 | 0.6159 | - |
162
- | 0.1491 | 2500 | 0.465 | - |
163
- | 0.1551 | 2600 | 1.0453 | - |
164
- | 0.1610 | 2700 | 0.565 | - |
165
- | 0.1670 | 2800 | 0.4328 | - |
166
- | 0.1729 | 2900 | 0.5229 | - |
167
- | 0.1789 | 3000 | 0.5581 | - |
168
- | 0.1849 | 3100 | 0.1847 | - |
169
- | 0.1908 | 3200 | 0.4755 | - |
170
- | 0.1968 | 3300 | 0.8408 | - |
171
- | 0.2028 | 3400 | 0.4852 | - |
172
- | 0.2087 | 3500 | 0.6054 | - |
173
- | 0.2147 | 3600 | 0.4868 | - |
174
- | 0.2207 | 3700 | 0.4138 | - |
175
- | 0.2266 | 3800 | 0.9303 | - |
176
- | 0.2326 | 3900 | 0.3892 | - |
177
- | 0.2385 | 4000 | 0.3462 | - |
178
- | 0.2445 | 4100 | 0.3579 | - |
179
- | 0.2505 | 4200 | 0.203 | - |
180
- | 0.2564 | 4300 | 0.4673 | - |
181
- | 0.2624 | 4400 | 0.1183 | - |
182
- | 0.2684 | 4500 | 0.506 | - |
183
- | 0.2743 | 4600 | 0.1378 | - |
184
- | 0.2803 | 4700 | 0.1603 | - |
185
- | 0.2863 | 4800 | 0.2337 | - |
186
- | 0.2922 | 4900 | 0.1526 | - |
187
- | 0.2982 | 5000 | 0.3597 | - |
188
- | 0.3042 | 5100 | 0.0672 | - |
189
- | 0.3101 | 5200 | 0.2134 | - |
190
- | 0.3161 | 5300 | 0.3521 | - |
191
- | 0.3220 | 5400 | 0.1098 | - |
192
- | 0.3280 | 5500 | 0.0723 | - |
193
- | 0.3340 | 5600 | 0.0349 | - |
194
- | 0.3399 | 5700 | 0.1389 | - |
195
- | 0.3459 | 5800 | 0.0966 | - |
196
- | 0.3519 | 5900 | 0.0998 | - |
197
- | 0.3578 | 6000 | 0.0263 | - |
198
- | 0.3638 | 6100 | 0.2343 | - |
199
- | 0.3698 | 6200 | 0.0776 | - |
200
- | 0.3757 | 6300 | 0.0037 | - |
201
- | 0.3817 | 6400 | 0.1324 | - |
202
- | 0.3876 | 6500 | 0.1259 | - |
203
- | 0.3936 | 6600 | 0.0197 | - |
204
- | 0.3996 | 6700 | 0.048 | - |
205
- | 0.4055 | 6800 | 0.077 | - |
206
- | 0.4115 | 6900 | 0.025 | - |
207
- | 0.4175 | 7000 | 0.1416 | - |
208
- | 0.4234 | 7100 | 0.0622 | - |
209
- | 0.4294 | 7200 | 0.0625 | - |
210
- | 0.4354 | 7300 | 0.0281 | - |
211
- | 0.4413 | 7400 | 0.0308 | - |
212
- | 0.4473 | 7500 | 0.0675 | - |
213
- | 0.4532 | 7600 | 0.0551 | - |
214
- | 0.4592 | 7700 | 0.0174 | - |
215
- | 0.4652 | 7800 | 0.0719 | - |
216
- | 0.4711 | 7900 | 0.0426 | - |
217
- | 0.4771 | 8000 | 0.0231 | - |
218
- | 0.4831 | 8100 | 0.0253 | - |
219
- | 0.4890 | 8200 | 0.0106 | - |
220
- | 0.4950 | 8300 | 0.0199 | - |
221
- | 0.5010 | 8400 | 0.0181 | - |
222
- | 0.5069 | 8500 | 0.0136 | - |
223
- | 0.5129 | 8600 | 0.0378 | - |
224
- | 0.5188 | 8700 | 0.0151 | - |
225
- | 0.5248 | 8800 | 0.002 | - |
226
- | 0.5308 | 8900 | 0.0008 | - |
227
- | 0.5367 | 9000 | 0.0025 | - |
228
- | 0.5427 | 9100 | 0.0125 | - |
229
- | 0.5487 | 9200 | 0.0112 | - |
230
- | 0.5546 | 9300 | 0.0019 | - |
231
- | 0.5606 | 9400 | 0.0265 | - |
232
- | 0.5666 | 9500 | 0.017 | - |
233
- | 0.5725 | 9600 | 0.0133 | - |
234
- | 0.5785 | 9700 | 0.0324 | - |
235
- | 0.5844 | 9800 | 0.0067 | - |
236
- | 0.5904 | 9900 | 0.0032 | - |
237
- | 0.5964 | 10000 | 0.0133 | - |
238
- | 0.6023 | 10100 | 0.0014 | - |
239
- | 0.6083 | 10200 | 0.0075 | - |
240
- | 0.6143 | 10300 | 0.0142 | - |
241
- | 0.6202 | 10400 | 0.0074 | - |
242
- | 0.6262 | 10500 | 0.0446 | - |
243
- | 0.6322 | 10600 | 0.0701 | - |
244
- | 0.6381 | 10700 | 0.0039 | - |
245
- | 0.6441 | 10800 | 0.0042 | - |
246
- | 0.6500 | 10900 | 0.004 | - |
247
- | 0.6560 | 11000 | 0.0009 | - |
248
- | 0.6620 | 11100 | 0.0007 | - |
249
- | 0.6679 | 11200 | 0.0012 | - |
250
- | 0.6739 | 11300 | 0.0178 | - |
251
- | 0.6799 | 11400 | 0.0024 | - |
252
- | 0.6858 | 11500 | 0.0006 | - |
253
- | 0.6918 | 11600 | 0.0011 | - |
254
- | 0.6978 | 11700 | 0.0043 | - |
255
- | 0.7037 | 11800 | 0.0013 | - |
256
- | 0.7097 | 11900 | 0.0019 | - |
257
- | 0.7156 | 12000 | 0.0025 | - |
258
- | 0.7216 | 12100 | 0.0004 | - |
259
- | 0.7276 | 12200 | 0.0065 | - |
260
- | 0.7335 | 12300 | 0.001 | - |
261
- | 0.7395 | 12400 | 0.0013 | - |
262
- | 0.7455 | 12500 | 0.0036 | - |
263
- | 0.7514 | 12600 | 0.0027 | - |
264
- | 0.7574 | 12700 | 0.0015 | - |
265
- | 0.7634 | 12800 | 0.0004 | - |
266
- | 0.7693 | 12900 | 0.0102 | - |
267
- | 0.7753 | 13000 | 0.0035 | - |
268
- | 0.7812 | 13100 | 0.0003 | - |
269
- | 0.7872 | 13200 | 0.0003 | - |
270
- | 0.7932 | 13300 | 0.0001 | - |
271
- | 0.7991 | 13400 | 0.0024 | - |
272
- | 0.8051 | 13500 | 0.0009 | - |
273
- | 0.8111 | 13600 | 0.0004 | - |
274
- | 0.8170 | 13700 | 0.0002 | - |
275
- | 0.8230 | 13800 | 0.0002 | - |
276
- | 0.8290 | 13900 | 0.0005 | - |
277
- | 0.8349 | 14000 | 0.0015 | - |
278
- | 0.8409 | 14100 | 0.0035 | - |
279
- | 0.8469 | 14200 | 0.0004 | - |
280
- | 0.8528 | 14300 | 0.0003 | - |
281
- | 0.8588 | 14400 | 0.0006 | - |
282
- | 0.8647 | 14500 | 0.0002 | - |
283
- | 0.8707 | 14600 | 0.0002 | - |
284
- | 0.8767 | 14700 | 0.0004 | - |
285
- | 0.8826 | 14800 | 0.0002 | - |
286
- | 0.8886 | 14900 | 0.0004 | - |
287
- | 0.8946 | 15000 | 0.0001 | - |
288
- | 0.9005 | 15100 | 0.0004 | - |
289
- | 0.9065 | 15200 | 0.0004 | - |
290
- | 0.9125 | 15300 | 0.0003 | - |
291
- | 0.9184 | 15400 | 0.0002 | - |
292
- | 0.9244 | 15500 | 0.0001 | - |
293
- | 0.9303 | 15600 | 0.0002 | - |
294
- | 0.9363 | 15700 | 0.0004 | - |
295
- | 0.9423 | 15800 | 0.0002 | - |
296
- | 0.9482 | 15900 | 0.0004 | - |
297
- | 0.9542 | 16000 | 0.0005 | - |
298
- | 0.9602 | 16100 | 0.0002 | - |
299
- | 0.9661 | 16200 | 0.0003 | - |
300
- | 0.9721 | 16300 | 0.0001 | - |
301
- | 0.9781 | 16400 | 0.0001 | - |
302
- | 0.9840 | 16500 | 0.0002 | - |
303
- | 0.9900 | 16600 | 0.0003 | - |
304
- | 0.9959 | 16700 | 0.0005 | - |
305
-
306
  ### Framework Versions
307
  - Python: 3.10.4
308
  - SetFit: 1.1.2
 
30
  # Bootstrapping a Sentence-Level Corpus Quality Classifier for Web Text using Active Learning (RANLP25)
31
 
32
  A multi-label sentence classifier trained with Active Learning for predicting high- or low-qality labels of german webtext.
33
+
34
  Training and evaluation code: <https://github.com/maximilian-bley/german-webtext-quality-classification>
35
 
36
  ## Model Details
37
 
38
+ **Labels**
39
 
40
  - **0=Sentence Boundary:** Sentence boundary errors occur if the start or ending of a sentence is malformed. This is the case if it begins with a lower case letter or an atypical character, or lacks a proper terminal punctuation mark (e.g., period, exclamation mark, or question mark).
41
 
 
112
 
113
  ## Training Details
114
 
 
115
  ### Training Hyperparameters
116
  - batch_size: (8, 8)
117
  - num_epochs: (1, 16)
 
131
  - eval_max_steps: -1
132
  - load_best_model_at_end: False
133
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
134
  ### Framework Versions
135
  - Python: 3.10.4
136
  - SetFit: 1.1.2