Hugging Face's logo

qanastek
/
pos-french-camembert-flair

Token Classification
Flair
PyTorch
French
sequence-tagger-model
Model card Files Community
qanastek commited on
Commit
d2013f6
·
1 Parent(s): b8b7fb0

Update

Browse files
Files changed (7) hide show
  1. README.md +63 -65
  2. dev.tsv +0 -0
  3. loss.tsv +116 -51
  4. pytorch_model.bin +2 -2
  5. test.tsv +0 -0
  6. train.py +54 -0
  7. training.log +0 -0
README.md CHANGED
@@ -11,9 +11,9 @@ widget:
11
  # POET: A French Extended Part-of-Speech Tagger
12
 
13
  - Corpora: [ANTILLES](https://github.com/qanastek/ANTILLES)
14
- - Embeddings: [Contextual String Embeddings for Sequence Labelling](https://aclanthology.org/C18-1139/) + [CamemBERT](https://arxiv.org/abs/1911.03894)
15
  - Sequence Labelling: [Bi-LSTM-CRF](https://arxiv.org/abs/1011.4088)
16
- - Number of Epochs: 50
17
 
18
  **People Involved**
19
 
@@ -34,7 +34,7 @@ from flair.data import Sentence
34
  from flair.models import SequenceTagger
35
 
36
  # Load the model
37
- model = SequenceTagger.load("qanastek/pos-french-camembert-flair")
38
 
39
  sentence = Sentence("George Washington est allé à Washington")
40
 
@@ -141,78 +141,76 @@ The test corpora used for this evaluation is available on [Github](https://githu
141
 
142
  ```plain
143
  Results:
144
- - F-score (micro) 0.9797
145
- - F-score (macro) 0.9178
146
- - Accuracy 0.9797
147
 
148
  By class:
149
  precision recall f1-score support
150
-
151
- PREP 0.9966 0.9987 0.9976 1483
152
- PUNCT 1.0000 1.0000 1.0000 833
153
- NMS 0.9634 0.9801 0.9717 753
154
- DET 0.9923 0.9984 0.9954 645
155
- VERB 0.9913 0.9811 0.9862 583
156
- NFS 0.9667 0.9839 0.9752 560
157
- ADV 0.9940 0.9821 0.9880 504
158
- PROPN 0.9541 0.8937 0.9229 395
159
- DETMS 1.0000 1.0000 1.0000 362
160
- AUX 0.9860 0.9915 0.9888 355
161
- YPFOR 1.0000 1.0000 1.0000 353
162
- NMP 0.9666 0.9475 0.9570 305
163
- COCO 0.9959 1.0000 0.9980 245
164
- ADJMS 0.9463 0.9385 0.9424 244
 
 
 
 
 
 
 
 
 
 
 
165
  DETFS 1.0000 1.0000 1.0000 240
166
- CHIF 0.9648 0.9865 0.9755 222
167
- NFP 0.9515 0.9849 0.9679 199
168
- ADJFS 0.9657 0.9286 0.9468 182
169
- VPPMS 0.9387 0.9745 0.9563 157
170
- COSUB 1.0000 0.9844 0.9921 128
171
- DINTMS 0.9918 0.9918 0.9918 122
172
- XFAMIL 0.9298 0.9217 0.9258 115
173
- PPER3MS 1.0000 1.0000 1.0000 87
174
- ADJMP 0.9294 0.9634 0.9461 82
175
- PDEMMS 1.0000 1.0000 1.0000 75
176
- ADJFP 0.9861 0.9342 0.9595 76
177
- PREL 0.9859 1.0000 0.9929 70
178
- DINTFS 0.9839 1.0000 0.9919 61
179
- PREF 1.0000 1.0000 1.0000 52
180
- PPOBJMS 0.9565 0.9362 0.9462 47
181
- PREFP 0.9778 1.0000 0.9888 44
182
- PINDMS 1.0000 0.9773 0.9885 44
183
- VPPFS 0.8298 0.9750 0.8966 40
184
- PPER1S 1.0000 1.0000 1.0000 42
185
- SYM 1.0000 0.9474 0.9730 38
186
- NOUN 0.8824 0.7692 0.8219 39
187
- PRON 1.0000 0.9677 0.9836 31
188
- PDEMFS 1.0000 1.0000 1.0000 29
189
- VPPMP 0.9286 1.0000 0.9630 26
190
- ADJ 0.9524 0.9091 0.9302 22
191
- PPER3MP 1.0000 1.0000 1.0000 20
192
- VPPFP 1.0000 1.0000 1.0000 19
193
- PPER3FS 1.0000 1.0000 1.0000 18
194
- MOTINC 0.3333 0.4000 0.3636 15
195
- PREFS 1.0000 1.0000 1.0000 10
196
- PPOBJMP 1.0000 0.8000 0.8889 10
197
- PPOBJFS 0.6250 0.8333 0.7143 6
198
- INTJ 0.5000 0.6667 0.5714 6
199
- PART 1.0000 1.0000 1.0000 4
200
- PDEMMP 1.0000 1.0000 1.0000 3
201
- PDEMFP 1.0000 1.0000 1.0000 3
202
  PPER3FP 1.0000 1.0000 1.0000 2
 
 
 
 
 
 
 
203
  NUM 1.0000 0.3333 0.5000 3
204
  PPER2S 1.0000 1.0000 1.0000 2
205
- PPOBJFP 0.5000 0.5000 0.5000 2
206
- PRELMS 1.0000 1.0000 1.0000 2
207
- PINDFS 0.5000 1.0000 0.6667 1
208
- PINDMP 1.0000 1.0000 1.0000 1
209
  X 0.0000 0.0000 0.0000 1
 
210
  PINDFP 1.0000 1.0000 1.0000 1
211
 
212
- micro avg 0.9797 0.9797 0.9797 10019
213
- macro avg 0.9228 0.9230 0.9178 10019
214
- weighted avg 0.9802 0.9797 0.9798 10019
215
- samples avg 0.9797 0.9797 0.9797 10019
216
  ```
217
 
218
  ## BibTeX Citations
 
11
  # POET: A French Extended Part-of-Speech Tagger
12
 
13
  - Corpora: [ANTILLES](https://github.com/qanastek/ANTILLES)
14
+ - Embeddings: [FastText](https://fasttext.cc/)
15
  - Sequence Labelling: [Bi-LSTM-CRF](https://arxiv.org/abs/1011.4088)
16
+ - Number of Epochs: 115
17
 
18
  **People Involved**
19
 
 
34
  from flair.models import SequenceTagger
35
 
36
  # Load the model
37
+ model = SequenceTagger.load("qanastek/pos-french")
38
 
39
  sentence = Sentence("George Washington est allé à Washington")
40
 
 
141
 
142
  ```plain
143
  Results:
144
+ - F-score (micro): 0.952
145
+ - F-score (macro): 0.8644
146
+ - Accuracy (incl. no class): 0.952
147
 
148
  By class:
149
  precision recall f1-score support
150
+ PPER1S 0.9767 1.0000 0.9882 42
151
+ VERB 0.9823 0.9537 0.9678 583
152
+ COSUB 0.9344 0.8906 0.9120 128
153
+ PUNCT 0.9878 0.9688 0.9782 833
154
+ PREP 0.9767 0.9879 0.9822 1483
155
+ PDEMMS 0.9583 0.9200 0.9388 75
156
+ COCO 0.9839 1.0000 0.9919 245
157
+ DET 0.9679 0.9814 0.9746 645
158
+ NMP 0.9521 0.9115 0.9313 305
159
+ ADJMP 0.8352 0.9268 0.8786 82
160
+ PREL 0.9324 0.9857 0.9583 70
161
+ PREFP 0.9767 0.9545 0.9655 44
162
+ AUX 0.9537 0.9859 0.9695 355
163
+ ADV 0.9440 0.9365 0.9402 504
164
+ VPPMP 0.8667 1.0000 0.9286 26
165
+ DINTMS 0.9919 1.0000 0.9959 122
166
+ ADJMS 0.9020 0.9057 0.9039 244
167
+ NMS 0.9226 0.9336 0.9281 753
168
+ NFS 0.9347 0.9714 0.9527 560
169
+ YPFOR 0.9806 1.0000 0.9902 353
170
+ PINDMS 1.0000 0.9091 0.9524 44
171
+ NOUN 0.8400 0.5385 0.6562 39
172
+ PROPN 0.8605 0.8278 0.8439 395
173
+ DETMS 0.9972 0.9972 0.9972 362
174
+ PPER3MS 0.9341 0.9770 0.9551 87
175
+ VPPMS 0.8994 0.9682 0.9325 157
176
  DETFS 1.0000 1.0000 1.0000 240
177
+ ADJFS 0.9266 0.9011 0.9136 182
178
+ ADJFP 0.9726 0.9342 0.9530 76
179
+ NFP 0.9463 0.9749 0.9604 199
180
+ VPPFS 0.8000 0.9000 0.8471 40
181
+ CHIF 0.9543 0.9414 0.9478 222
182
+ XFAMIL 0.9346 0.8696 0.9009 115
183
+ PPER3MP 0.9474 0.9000 0.9231 20
184
+ PPOBJMS 0.8800 0.9362 0.9072 47
185
+ PREF 0.8889 0.9231 0.9057 52
186
+ PPOBJMP 1.0000 0.6000 0.7500 10
187
+ SYM 0.9706 0.8684 0.9167 38
188
+ DINTFS 0.9683 1.0000 0.9839 61
189
+ PDEMFS 1.0000 0.8966 0.9455 29
190
+ PPER3FS 1.0000 0.9444 0.9714 18
191
+ VPPFP 0.9500 1.0000 0.9744 19
192
+ PRON 0.9200 0.7419 0.8214 31
193
+ PPOBJFS 0.8333 0.8333 0.8333 6
194
+ PART 0.8000 1.0000 0.8889 4
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
195
  PPER3FP 1.0000 1.0000 1.0000 2
196
+ MOTINC 0.3571 0.3333 0.3448 15
197
+ PDEMMP 1.0000 0.6667 0.8000 3
198
+ INTJ 0.4000 0.6667 0.5000 6
199
+ PREFS 1.0000 0.5000 0.6667 10
200
+ ADJ 0.7917 0.8636 0.8261 22
201
+ PINDMP 0.0000 0.0000 0.0000 1
202
+ PINDFS 1.0000 1.0000 1.0000 1
203
  NUM 1.0000 0.3333 0.5000 3
204
  PPER2S 1.0000 1.0000 1.0000 2
205
+ PPOBJFP 1.0000 0.5000 0.6667 2
206
+ PDEMFP 1.0000 0.6667 0.8000 3
 
 
207
  X 0.0000 0.0000 0.0000 1
208
+ PRELMS 1.0000 1.0000 1.0000 2
209
  PINDFP 1.0000 1.0000 1.0000 1
210
 
211
+ accuracy 0.9520 10019
212
+ macro avg 0.8956 0.8521 0.8644 10019
213
+ weighted avg 0.9524 0.9520 0.9515 10019
 
214
  ```
215
 
216
  ## BibTeX Citations
dev.tsv DELETED
The diff for this file is too large to render. See raw diff
 
loss.tsv CHANGED
@@ -1,51 +1,116 @@
1
- EPOCH TIMESTAMP BAD_EPOCHS LEARNING_RATE TRAIN_LOSS DEV_LOSS DEV_PRECISION DEV_RECALL DEV_F1 DEV_ACCURACY
2
- 1 08:40:38 0 0.1000 0.4139037357786823 0.09867297857999802 0.9723 0.9723 0.9723 0.9723
3
- 2 08:43:59 0 0.1000 0.18683743789460058 0.08219591528177261 0.9761 0.9761 0.9761 0.9761
4
- 3 08:47:19 0 0.1000 0.15330191376146995 0.07821641117334366 0.9771 0.9771 0.9771 0.9771
5
- 4 08:50:41 0 0.1000 0.13679069024466697 0.07048774510622025 0.9784 0.9784 0.9784 0.9784
6
- 5 08:54:05 0 0.1000 0.12696925415918978 0.06857253611087799 0.9795 0.9795 0.9795 0.9795
7
- 6 08:57:27 0 0.1000 0.11950518270160057 0.06588418781757355 0.9805 0.9805 0.9805 0.9805
8
- 7 09:00:50 0 0.1000 0.11493925645982497 0.06450950354337692 0.981 0.981 0.981 0.981
9
- 8 09:04:17 1 0.1000 0.10857167749940388 0.06390747427940369 0.9805 0.9805 0.9805 0.9805
10
- 9 09:07:32 0 0.1000 0.10607049356155288 0.06607701629400253 0.9814 0.9814 0.9814 0.9814
11
- 10 09:10:55 1 0.1000 0.10357679251794805 0.06536506861448288 0.9811 0.9811 0.9811 0.9811
12
- 11 09:14:14 2 0.1000 0.09934903912638607 0.06659943610429764 0.9811 0.9811 0.9811 0.9811
13
- 12 09:17:31 0 0.1000 0.09835840644128474 0.06410104781389236 0.9816 0.9816 0.9816 0.9816
14
- 13 09:20:54 1 0.1000 0.09667944757963298 0.06427688896656036 0.9816 0.9816 0.9816 0.9816
15
- 14 09:24:13 0 0.1000 0.09310611423129937 0.06639766693115234 0.9817 0.9817 0.9817 0.9817
16
- 15 09:27:36 0 0.1000 0.09273757020644302 0.06283392012119293 0.982 0.982 0.982 0.982
17
- 16 09:30:58 1 0.1000 0.0906242242911817 0.06354553997516632 0.982 0.982 0.982 0.982
18
- 17 09:34:16 0 0.1000 0.08953603486075622 0.06361010670661926 0.9823 0.9823 0.9823 0.9823
19
- 18 09:37:38 1 0.1000 0.08716175395396096 0.06376409530639648 0.982 0.982 0.982 0.982
20
- 19 09:40:54 0 0.1000 0.08630291839682559 0.06360483914613724 0.9824 0.9824 0.9824 0.9824
21
- 20 09:44:18 1 0.1000 0.08576645481590557 0.06494450569152832 0.982 0.982 0.982 0.982
22
- 21 09:47:35 0 0.1000 0.08385216420677111 0.06328344345092773 0.9827 0.9827 0.9827 0.9827
23
- 22 09:50:58 1 0.1000 0.08442341949455835 0.06346500664949417 0.9815 0.9815 0.9815 0.9815
24
- 23 09:54:16 2 0.1000 0.08142570236260006 0.06540019810199738 0.9821 0.9821 0.9821 0.9821
25
- 24 09:57:35 3 0.1000 0.0822403573790078 0.06453310698270798 0.9819 0.9819 0.9819 0.9819
26
- 25 10:00:51 4 0.1000 0.08115838012320148 0.06579063087701797 0.9817 0.9817 0.9817 0.9817
27
- 26 10:04:08 1 0.0500 0.07444606900847728 0.06646668165922165 0.9822 0.9822 0.9822 0.9822
28
- 27 10:07:27 2 0.0500 0.0712278272039567 0.06514652073383331 0.9823 0.9823 0.9823 0.9823
29
- 28 10:10:43 0 0.0500 0.07007554484263678 0.06285692006349564 0.9828 0.9828 0.9828 0.9828
30
- 29 10:14:05 0 0.0500 0.06775021975879568 0.06288447976112366 0.9831 0.9831 0.9831 0.9831
31
- 30 10:17:28 1 0.0500 0.06664810656288497 0.06311798095703125 0.9824 0.9824 0.9824 0.9824
32
- 31 10:20:44 2 0.0500 0.06655944385465427 0.06285466253757477 0.9829 0.9829 0.9829 0.9829
33
- 32 10:24:00 3 0.0500 0.06484422466931324 0.062373436987400055 0.9827 0.9827 0.9827 0.9827
34
- 33 10:27:18 4 0.0500 0.0640099294991078 0.06352584064006805 0.983 0.983 0.983 0.983
35
- 34 10:30:35 0 0.0250 0.060174371477019914 0.06348917633295059 0.9835 0.9835 0.9835 0.9835
36
- 35 10:33:57 1 0.0250 0.06001775798271323 0.06338120251893997 0.9829 0.9829 0.9829 0.9829
37
- 36 10:37:17 2 0.0250 0.05871860721139249 0.06424003839492798 0.9835 0.9835 0.9835 0.9835
38
- 37 10:40:35 3 0.0250 0.058616780999994414 0.06326954811811447 0.9831 0.9831 0.9831 0.9831
39
- 38 10:43:52 4 0.0250 0.05700768598441678 0.06343492120504379 0.9831 0.9831 0.9831 0.9831
40
- 39 10:47:11 1 0.0125 0.05521906535375693 0.06419230252504349 0.9829 0.9829 0.9829 0.9829
41
- 40 10:50:26 2 0.0125 0.0545923078244546 0.06343018263578415 0.9829 0.9829 0.9829 0.9829
42
- 41 10:53:43 3 0.0125 0.05370077676597374 0.06420625746250153 0.9831 0.9831 0.9831 0.9831
43
- 42 10:57:03 4 0.0125 0.053218689249659105 0.06362675130367279 0.9831 0.9831 0.9831 0.9831
44
- 43 11:00:20 1 0.0063 0.052319793331815405 0.06449297815561295 0.983 0.983 0.983 0.983
45
- 44 11:03:36 2 0.0063 0.052714465782813934 0.06455685943365097 0.9831 0.9831 0.9831 0.9831
46
- 45 11:06:56 3 0.0063 0.052344465870703995 0.06413871794939041 0.983 0.983 0.983 0.983
47
- 46 11:10:13 4 0.0063 0.05212448329383334 0.0644669309258461 0.983 0.983 0.983 0.983
48
- 47 11:13:29 1 0.0031 0.05113414183996207 0.06470787525177002 0.9829 0.9829 0.9829 0.9829
49
- 48 11:16:48 2 0.0031 0.05100923420506556 0.06484530121088028 0.983 0.983 0.983 0.983
50
- 49 11:20:04 3 0.0031 0.05102771810317176 0.06486314535140991 0.983 0.983 0.983 0.983
51
- 50 11:23:20 4 0.0031 0.05190557099563212 0.06452730298042297 0.9831 0.9831 0.9831 0.9831
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ EPOCH TIMESTAMP BAD_EPOCHS LEARNING_RATE TRAIN_LOSS DEV_LOSS DEV_ACCURACY
2
+ 1 14:02:20 0 0.1000 68.69758622836223 33.562286376953125 0.6341
3
+ 2 14:03:39 0 0.1000 28.112464617838903 14.451913833618164 0.842
4
+ 3 14:04:57 0 0.1000 18.645767338507998 9.90114974975586 0.8889
5
+ 4 14:06:15 0 0.1000 15.248240833788847 7.970705986022949 0.9102
6
+ 5 14:07:32 0 0.1000 13.478108287912555 7.087363243103027 0.9182
7
+ 6 14:08:49 0 0.1000 12.370107963021878 6.438025951385498 0.9242
8
+ 7 14:10:06 0 0.1000 11.615934616696519 5.967233657836914 0.9301
9
+ 8 14:11:23 0 0.1000 10.971208243243462 5.777602195739746 0.9305
10
+ 9 14:12:41 0 0.1000 10.598726795837942 5.393132209777832 0.9343
11
+ 10 14:13:57 1 0.1000 10.203757547699245 5.497661590576172 0.9333
12
+ 11 14:14:57 0 0.1000 9.893801364223513 5.045197486877441 0.9384
13
+ 12 14:16:14 0 0.1000 9.663424407486367 4.956379413604736 0.94
14
+ 13 14:17:34 0 0.1000 9.477814227078868 4.854061126708984 0.9407
15
+ 14 14:18:50 0 0.1000 9.267724408512622 4.6637468338012695 0.9437
16
+ 15 14:20:07 1 0.1000 9.154714643427756 4.723843574523926 0.9425
17
+ 16 14:21:03 0 0.1000 8.99992180714565 4.600645542144775 0.9449
18
+ 17 14:22:23 0 0.1000 8.837092483993125 4.431546211242676 0.9457
19
+ 18 14:23:44 0 0.1000 8.720927441014652 4.406030178070068 0.9458
20
+ 19 14:25:02 0 0.1000 8.62262540580952 4.378823757171631 0.9467
21
+ 20 14:26:19 0 0.1000 8.506264420737207 4.324984550476074 0.9473
22
+ 21 14:27:36 0 0.1000 8.41792546331355 4.271585941314697 0.9478
23
+ 22 14:28:57 0 0.1000 8.313587416589787 4.242696285247803 0.9481
24
+ 23 14:30:23 0 0.1000 8.193923979733897 4.192256927490234 0.9487
25
+ 24 14:31:43 0 0.1000 8.11061666708077 4.094086647033691 0.9502
26
+ 25 14:33:00 1 0.1000 7.994365789194022 4.010470390319824 0.9498
27
+ 26 14:34:00 0 0.1000 8.008020202670478 4.04341983795166 0.9504
28
+ 27 14:35:18 0 0.1000 7.865361327618624 3.950613498687744 0.9514
29
+ 28 14:36:35 1 0.1000 7.784816885416487 3.975137233734131 0.9506
30
+ 29 14:37:32 2 0.1000 7.764582988435188 3.91574764251709 0.9511
31
+ 30 14:38:28 0 0.1000 7.675331065085082 3.923271894454956 0.9516
32
+ 31 14:39:47 0 0.1000 7.6415734966244315 3.890695333480835 0.9531
33
+ 32 14:41:07 1 0.1000 7.5645937328845 3.8633809089660645 0.9521
34
+ 33 14:42:03 0 0.1000 7.523175163606627 3.8121325969696045 0.9532
35
+ 34 14:43:22 1 0.1000 7.426563718677622 3.7572150230407715 0.9526
36
+ 35 14:44:22 0 0.1000 7.3233047375636815 3.7696845531463623 0.9537
37
+ 36 14:45:39 1 0.1000 7.389569210795175 3.7234673500061035 0.953
38
+ 37 14:46:36 2 0.1000 7.281651627700941 3.734175443649292 0.9533
39
+ 38 14:47:32 3 0.1000 7.270007542804279 3.769817352294922 0.9531
40
+ 39 14:48:31 0 0.1000 7.213365968349761 3.6631696224212646 0.9543
41
+ 40 14:49:47 1 0.1000 7.149712216537611 3.682539939880371 0.9543
42
+ 41 14:50:44 2 0.1000 7.1777994527226 3.6655328273773193 0.954
43
+ 42 14:51:39 0 0.1000 7.042381501830785 3.650456428527832 0.9549
44
+ 43 14:52:56 0 0.1000 7.088383231542807 3.5950212478637695 0.9553
45
+ 44 14:54:15 1 0.1000 7.041419814118242 3.6201629638671875 0.955
46
+ 45 14:55:11 2 0.1000 6.940254409756281 3.6683692932128906 0.9547
47
+ 46 14:56:08 3 0.1000 6.920480015003576 3.6241424083709717 0.9546
48
+ 47 14:57:05 0 0.1000 6.890594191255823 3.5482797622680664 0.9559
49
+ 48 14:58:25 1 0.1000 6.855043533629021 3.5812222957611084 0.9553
50
+ 49 14:59:22 2 0.1000 6.800135021716093 3.53963565826416 0.9557
51
+ 50 15:00:18 0 0.1000 6.768724057526715 3.5272233486175537 0.9568
52
+ 51 15:01:36 1 0.1000 6.739383368365533 3.5689055919647217 0.9562
53
+ 52 15:02:33 2 0.1000 6.661376100725832 3.5081849098205566 0.9564
54
+ 53 15:03:30 3 0.1000 6.624162015661729 3.5275826454162598 0.9556
55
+ 54 15:04:27 4 0.1000 6.6407164852176095 3.4983420372009277 0.9567
56
+ 55 15:05:23 0 0.0500 6.55794746685872 3.457058906555176 0.9568
57
+ 56 15:06:41 0 0.0500 6.554000040071201 3.446054458618164 0.9568
58
+ 57 15:08:01 0 0.0500 6.505894108156188 3.4272050857543945 0.9569
59
+ 58 15:09:19 0 0.0500 6.536830239591345 3.4196906089782715 0.9577
60
+ 59 15:10:37 0 0.0500 6.450945086183801 3.4158873558044434 0.958
61
+ 60 15:11:58 1 0.0500 6.467335283228781 3.4086313247680664 0.9578
62
+ 61 15:12:54 2 0.0500 6.395290446492423 3.407853603363037 0.9575
63
+ 62 15:13:50 3 0.0500 6.424222988364971 3.4143223762512207 0.9575
64
+ 63 15:14:46 4 0.0500 6.408897661529811 3.399538516998291 0.9577
65
+ 64 15:15:43 0 0.0250 6.38559970602525 3.399369239807129 0.958
66
+ 65 15:17:01 1 0.0250 6.412552774479959 3.3750159740448 0.9576
67
+ 66 15:17:57 0 0.0250 6.337497213245493 3.3704581260681152 0.9581
68
+ 67 15:19:15 1 0.0250 6.324488298027917 3.38270902633667 0.958
69
+ 68 15:20:12 2 0.0250 6.331536335227764 3.3832640647888184 0.9581
70
+ 69 15:21:08 3 0.0250 6.313072968373256 3.3663582801818848 0.958
71
+ 70 15:22:04 0 0.0250 6.234962488697693 3.3701589107513428 0.9581
72
+ 71 15:23:23 0 0.0250 6.320951318318865 3.3578968048095703 0.9581
73
+ 72 15:24:43 1 0.0250 6.296895212831751 3.358468770980835 0.9579
74
+ 73 15:25:40 2 0.0250 6.242309089255544 3.3546149730682373 0.9579
75
+ 74 15:26:37 0 0.0250 6.253713097192545 3.3542890548706055 0.9582
76
+ 75 15:27:53 0 0.0250 6.227823793360617 3.352653980255127 0.9585
77
+ 76 15:29:09 1 0.0250 6.255534311311435 3.3581032752990723 0.9581
78
+ 77 15:30:06 2 0.0250 6.240572617117283 3.3542392253875732 0.9578
79
+ 78 15:31:03 3 0.0250 6.261245811935019 3.3449277877807617 0.9583
80
+ 79 15:31:59 4 0.0250 6.227426613326621 3.348515510559082 0.9584
81
+ 80 15:32:56 0 0.0125 6.214050103077846 3.3447864055633545 0.9585
82
+ 81 15:34:13 1 0.0125 6.221008515990941 3.3493881225585938 0.9581
83
+ 82 15:35:10 2 0.0125 6.221904636484332 3.3341634273529053 0.9583
84
+ 83 15:36:06 3 0.0125 6.160273648996268 3.338702917098999 0.9583
85
+ 84 15:37:06 4 0.0125 6.138500496349503 3.328439235687256 0.9583
86
+ 85 15:38:03 1 0.0063 6.161158080649587 3.3330752849578857 0.9582
87
+ 86 15:39:00 2 0.0063 6.151211350365022 3.3295114040374756 0.9583
88
+ 87 15:39:57 3 0.0063 6.232518280501914 3.32422137260437 0.9583
89
+ 88 15:40:54 4 0.0063 6.236182183291005 3.3302392959594727 0.9583
90
+ 89 15:41:51 0 0.0031 6.141926377220491 3.324725866317749 0.9585
91
+ 90 15:43:08 1 0.0031 6.166836645750873 3.3286867141723633 0.9583
92
+ 91 15:44:05 2 0.0031 6.13570765689411 3.3274388313293457 0.9582
93
+ 92 15:45:02 3 0.0031 6.138865977261974 3.3267321586608887 0.9583
94
+ 93 15:45:59 4 0.0031 6.161072790095236 3.3274779319763184 0.9583
95
+ 94 15:46:55 1 0.0016 6.144343996469954 3.3262553215026855 0.9583
96
+ 95 15:47:52 2 0.0016 6.172166739944863 3.3275344371795654 0.9583
97
+ 96 15:48:52 3 0.0016 6.184053657329188 3.326869487762451 0.9583
98
+ 97 15:49:49 4 0.0016 6.170179907199556 3.330195903778076 0.9585
99
+ 98 15:50:45 1 0.0008 6.126421856669198 3.3274950981140137 0.9583
100
+ 99 15:51:42 2 0.0008 6.152514440823445 3.3275747299194336 0.9584
101
+ 100 15:52:38 0 0.0008 6.157003947063885 3.327719211578369 0.9586
102
+ 101 15:53:55 1 0.0008 6.169898602814801 3.327259063720703 0.9585
103
+ 102 15:54:52 2 0.0008 6.145160662389435 3.325741767883301 0.9585
104
+ 103 15:55:48 0 0.0008 6.15242785901095 3.325845241546631 0.9586
105
+ 104 15:57:06 1 0.0008 6.170637012582965 3.3250930309295654 0.9585
106
+ 105 15:58:03 2 0.0008 6.101018842342681 3.32464861869812 0.9585
107
+ 106 15:58:59 3 0.0008 6.18395307844719 3.3257997035980225 0.9585
108
+ 107 15:59:56 4 0.0008 6.165698562048178 3.3272788524627686 0.9585
109
+ 108 16:00:52 1 0.0004 6.123404840452481 3.32755184173584 0.9586
110
+ 109 16:01:52 2 0.0004 6.168317748382028 3.3270320892333984 0.9585
111
+ 110 16:02:49 3 0.0004 6.15737580831072 3.3262462615966797 0.9584
112
+ 111 16:03:45 4 0.0004 6.181477074074534 3.325368881225586 0.9584
113
+ 112 16:04:42 1 0.0002 6.151819916953028 3.3254475593566895 0.9585
114
+ 113 16:05:39 2 0.0002 6.152685663341421 3.325159788131714 0.9585
115
+ 114 16:06:36 3 0.0002 6.114718158688166 3.325321912765503 0.9584
116
+ 115 16:07:33 4 0.0002 6.157736689643523 3.3252782821655273 0.9585
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:81250a3c3c5e6639a92f0a5c782ca1533c886a4ae9d48dccf7fffb85c8c27794
3
- size 539091093
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:58943a5be6d663da905aa8238e84d839aaef209dfec8ce716ed1ffe5e137cdee
3
+ size 1245112805
test.tsv CHANGED
The diff for this file is too large to render. See raw diff
 
train.py ADDED
@@ -0,0 +1,54 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import os
2
+ import argparse
3
+ from datetime import datetime
4
+
5
+ from flair.data import Corpus
6
+ from flair.models import SequenceTagger
7
+ from flair.trainers import ModelTrainer
8
+ from flair.datasets import UniversalDependenciesCorpus
9
+ from flair.embeddings import WordEmbeddings, StackedEmbeddings
10
+
11
+ parser = argparse.ArgumentParser(description='Flair Training Part-of-speech tagging')
12
+ parser.add_argument('-output', type=str, default="models/", help='The output directory')
13
+ parser.add_argument('-epochs', type=int, default=1, help='Number of Epochs')
14
+ args = parser.parse_args()
15
+
16
+ output = os.path.join(args.output, "UPOS_UD_FRENCH_PLUS_" + str(args.epochs) + "_" + datetime.today().strftime('%Y-%m-%d-%H:%M:%S'))
17
+ print(output)
18
+
19
+ # corpus: Corpus = UD_FRENCH()
20
+ corpus: Corpus = UniversalDependenciesCorpus(
21
+ data_folder='UD_FRENCH_PLUS',
22
+ train_file="fr_gsd-ud-train.conllu",
23
+ test_file="fr_gsd-ud-test.conllu",
24
+ dev_file="fr_gsd-ud-dev.conllu",
25
+ )
26
+ # print(corpus)
27
+
28
+ tag_type = 'upos'
29
+
30
+ tag_dictionary = corpus.make_tag_dictionary(tag_type=tag_type)
31
+ # print(tag_dictionary)
32
+
33
+ embedding_types = [
34
+ WordEmbeddings('fr'),
35
+ ]
36
+
37
+ embeddings: StackedEmbeddings = StackedEmbeddings(embeddings=embedding_types)
38
+
39
+ tagger: SequenceTagger = SequenceTagger(
40
+ hidden_size=256,
41
+ embeddings=embeddings,
42
+ tag_dictionary=tag_dictionary,
43
+ tag_type=tag_type,
44
+ use_crf=True
45
+ )
46
+
47
+ trainer: ModelTrainer = ModelTrainer(tagger, corpus)
48
+
49
+ trainer.train(
50
+ output,
51
+ learning_rate=0.1,
52
+ mini_batch_size=128,
53
+ max_epochs=args.epochs
54
+ )
training.log CHANGED
The diff for this file is too large to render. See raw diff