Hugging Face's logo

oliverguhr
/
spelling-correction-german-base

Text2Text Generation
Transformers
PyTorch
TensorBoard
Safetensors
t5
Generated from Trainer
text-generation-inference
Inference Endpoints
Model card Files Metrics Community
3
oliverguhr commited on
Commit
d347a22
•
1 Parent(s): 370b45c

added new flan model

Browse files
Files changed (1) hide show
  1. README.md +198 -14
README.md CHANGED
@@ -3,31 +3,215 @@ license: apache-2.0
3
  tags:
4
  - generated_from_trainer
5
  model-index:
6
- - name: bart-base-spelling-de
7
  results: []
8
- widget:
9
- - text: "das idst ein neuZr test"
10
- example_title: "1"
11
  ---
12
 
13
- This is an experimental model that should fix your typos and punctuation.
14
- If you like to run your own experiments or train for a different language, have a look at [the code](https://github.com/oliverguhr/spelling).
15
 
 
 
 
 
 
 
16
 
17
  ## Model description
18
 
19
- This is a proof of concept spelling correction model for german.
20
 
21
  ## Intended uses & limitations
22
 
23
- This is a work in progress, be aware that the model can produce artefacts.
24
- You can test the model using the pipeline-interface:
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
25
 
26
- ```python
27
- from transformers import pipeline
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
28
 
29
- fix_spelling = pipeline("text2text-generation",model="oliverguhr/spelling-correction-german-base")
30
 
31
- print(fix_spelling("das idst ein neuZr test",max_length=2048))
32
- ```
33
 
 
 
 
 
 
3
  tags:
4
  - generated_from_trainer
5
  model-index:
6
+ - name: flan-t5-spelling-de-base-fullds2
7
  results: []
 
 
 
8
  ---
9
 
10
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
11
+ should probably proofread and complete it, then remove this comment. -->
12
 
13
+ # flan-t5-spelling-de-base-fullds2
14
+
15
+ This model is a fine-tuned version of [google/flan-t5-base](https://huggingface.co/google/flan-t5-base) on an unknown dataset.
16
+ It achieves the following results on the evaluation set:
17
+ - Loss: 0.0525
18
+ - Cer: 0.0077
19
 
20
  ## Model description
21
 
22
+ More information needed
23
 
24
  ## Intended uses & limitations
25
 
26
+ More information needed
27
+
28
+ ## Training and evaluation data
29
+
30
+ More information needed
31
+
32
+ ## Training procedure
33
+
34
+ ### Training hyperparameters
35
+
36
+ The following hyperparameters were used during training:
37
+ - learning_rate: 0.003
38
+ - train_batch_size: 16
39
+ - eval_batch_size: 16
40
+ - seed: 42
41
+ - gradient_accumulation_steps: 16
42
+ - total_train_batch_size: 256
43
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
44
+ - lr_scheduler_type: linear
45
+ - num_epochs: 2.0
46
+
47
+ ### Training results
48
 
49
+ | Training Loss | Epoch | Step | Validation Loss | Cer |
50
+ |:-------------:|:-----:|:-----:|:---------------:|:------:|
51
+ | 0.444 | 0.01 | 500 | 0.3389 | 0.6733 |
52
+ | 0.3429 | 0.03 | 1000 | 0.2655 | 0.6706 |
53
+ | 0.2914 | 0.04 | 1500 | 0.2277 | 0.6705 |
54
+ | 0.264 | 0.05 | 2000 | 0.2078 | 0.6698 |
55
+ | 0.2506 | 0.06 | 2500 | 0.1894 | 0.6694 |
56
+ | 0.2305 | 0.08 | 3000 | 0.1787 | 0.6685 |
57
+ | 0.2206 | 0.09 | 3500 | 0.1685 | 0.6688 |
58
+ | 0.2086 | 0.1 | 4000 | 0.1607 | 0.6685 |
59
+ | 0.1955 | 0.11 | 4500 | 0.1518 | 0.6683 |
60
+ | 0.1903 | 0.13 | 5000 | 0.1475 | 0.6686 |
61
+ | 0.1827 | 0.14 | 5500 | 0.1430 | 0.6684 |
62
+ | 0.1775 | 0.15 | 6000 | 0.1369 | 0.6681 |
63
+ | 0.1748 | 0.16 | 6500 | 0.1359 | 0.6681 |
64
+ | 0.1725 | 0.18 | 7000 | 0.1312 | 0.6677 |
65
+ | 0.1638 | 0.19 | 7500 | 0.1254 | 0.6674 |
66
+ | 0.1575 | 0.2 | 8000 | 0.1255 | 0.6680 |
67
+ | 0.1537 | 0.21 | 8500 | 0.1204 | 0.6678 |
68
+ | 0.1516 | 0.23 | 9000 | 0.1188 | 0.6671 |
69
+ | 0.1526 | 0.24 | 9500 | 0.1150 | 0.6673 |
70
+ | 0.148 | 0.25 | 10000 | 0.1131 | 0.6676 |
71
+ | 0.1445 | 0.26 | 10500 | 0.1107 | 0.6675 |
72
+ | 0.1378 | 0.28 | 11000 | 0.1113 | 0.6664 |
73
+ | 6.1099 | 0.29 | 11500 | 6.0484 | 0.8805 |
74
+ | 4.9528 | 0.3 | 12000 | 4.6614 | 0.8115 |
75
+ | 0.2066 | 0.31 | 12500 | 0.1495 | 0.6679 |
76
+ | 0.1654 | 0.33 | 13000 | 0.1228 | 0.6678 |
77
+ | 0.1552 | 0.34 | 13500 | 0.1153 | 0.6670 |
78
+ | 0.1443 | 0.35 | 14000 | 0.1110 | 0.6670 |
79
+ | 0.1397 | 0.36 | 14500 | 0.1073 | 0.6670 |
80
+ | 0.1366 | 0.38 | 15000 | 0.1067 | 0.6664 |
81
+ | 0.1362 | 0.39 | 15500 | 0.1043 | 0.6669 |
82
+ | 0.1375 | 0.4 | 16000 | 0.1012 | 0.6668 |
83
+ | 0.1325 | 0.41 | 16500 | 0.0996 | 0.6672 |
84
+ | 0.1277 | 0.43 | 17000 | 0.0993 | 0.6664 |
85
+ | 0.1261 | 0.44 | 17500 | 0.0977 | 0.6667 |
86
+ | 0.1274 | 0.45 | 18000 | 0.0978 | 0.6666 |
87
+ | 0.127 | 0.46 | 18500 | 0.0952 | 0.6670 |
88
+ | 0.1218 | 0.48 | 19000 | 0.0933 | 0.6666 |
89
+ | 0.1196 | 0.49 | 19500 | 0.0923 | 0.6670 |
90
+ | 0.1192 | 0.5 | 20000 | 0.0920 | 0.6665 |
91
+ | 0.1171 | 0.52 | 20500 | 0.0910 | 0.6664 |
92
+ | 0.1153 | 0.53 | 21000 | 0.0906 | 0.6667 |
93
+ | 0.1102 | 0.54 | 21500 | 0.0890 | 0.6669 |
94
+ | 0.1147 | 0.55 | 22000 | 0.0886 | 0.6667 |
95
+ | 0.1144 | 0.57 | 22500 | 0.0868 | 0.6664 |
96
+ | 0.1132 | 0.58 | 23000 | 0.0858 | 0.6666 |
97
+ | 0.1073 | 0.59 | 23500 | 0.0853 | 0.6667 |
98
+ | 0.109 | 0.6 | 24000 | 0.0845 | 0.6663 |
99
+ | 0.1073 | 0.62 | 24500 | 0.0842 | 0.6662 |
100
+ | 0.1062 | 0.63 | 25000 | 0.0831 | 0.6662 |
101
+ | 0.1018 | 0.64 | 25500 | 0.0830 | 0.6662 |
102
+ | 0.1052 | 0.65 | 26000 | 0.0818 | 0.6666 |
103
+ | 0.1072 | 0.67 | 26500 | 0.0811 | 0.6662 |
104
+ | 0.1023 | 0.68 | 27000 | 0.0807 | 0.6661 |
105
+ | 0.1013 | 0.69 | 27500 | 0.0801 | 0.6664 |
106
+ | 0.0986 | 0.7 | 28000 | 0.0797 | 0.6664 |
107
+ | 0.1022 | 0.72 | 28500 | 0.0786 | 0.6662 |
108
+ | 0.0984 | 0.73 | 29000 | 0.0781 | 0.6659 |
109
+ | 0.0971 | 0.74 | 29500 | 0.0778 | 0.6662 |
110
+ | 0.0963 | 0.75 | 30000 | 0.0773 | 0.6660 |
111
+ | 0.0958 | 0.77 | 30500 | 0.0760 | 0.6662 |
112
+ | 0.0999 | 0.78 | 31000 | 0.0760 | 0.6661 |
113
+ | 0.0953 | 0.79 | 31500 | 0.0752 | 0.6661 |
114
+ | 0.095 | 0.8 | 32000 | 0.0749 | 0.6662 |
115
+ | 0.09 | 0.82 | 32500 | 0.0748 | 0.6663 |
116
+ | 0.0927 | 0.83 | 33000 | 0.0740 | 0.6656 |
117
+ | 0.0914 | 0.84 | 33500 | 0.0739 | 0.6662 |
118
+ | 0.0889 | 0.85 | 34000 | 0.0737 | 0.6659 |
119
+ | 0.0924 | 0.87 | 34500 | 0.0726 | 0.6660 |
120
+ | 0.0898 | 0.88 | 35000 | 0.0719 | 0.6659 |
121
+ | 0.0913 | 0.89 | 35500 | 0.0721 | 0.6657 |
122
+ | 0.0897 | 0.9 | 36000 | 0.0715 | 0.6657 |
123
+ | 0.0887 | 0.92 | 36500 | 0.0708 | 0.6659 |
124
+ | 0.0922 | 0.93 | 37000 | 0.0712 | 0.6653 |
125
+ | 0.0905 | 0.94 | 37500 | 0.0707 | 0.6660 |
126
+ | 0.0881 | 0.95 | 38000 | 0.0700 | 0.6658 |
127
+ | 0.0858 | 0.97 | 38500 | 0.0693 | 0.6658 |
128
+ | 0.0882 | 0.98 | 39000 | 0.0690 | 0.6657 |
129
+ | 0.0858 | 0.99 | 39500 | 0.0688 | 0.6656 |
130
+ | 0.0808 | 1.0 | 40000 | 0.0680 | 0.6658 |
131
+ | 0.0783 | 1.02 | 40500 | 0.0680 | 0.6657 |
132
+ | 0.0822 | 1.03 | 41000 | 0.0676 | 0.6658 |
133
+ | 0.077 | 1.04 | 41500 | 0.0675 | 0.6657 |
134
+ | 0.0788 | 1.06 | 42000 | 0.0673 | 0.6655 |
135
+ | 0.0754 | 1.07 | 42500 | 0.0667 | 0.6660 |
136
+ | 0.0762 | 1.08 | 43000 | 0.0669 | 0.6656 |
137
+ | 0.075 | 1.09 | 43500 | 0.0660 | 0.6660 |
138
+ | 0.0816 | 1.11 | 44000 | 0.0661 | 0.6657 |
139
+ | 0.0758 | 1.12 | 44500 | 0.0659 | 0.6657 |
140
+ | 0.0767 | 1.13 | 45000 | 0.0653 | 0.6658 |
141
+ | 0.076 | 1.14 | 45500 | 0.0649 | 0.6656 |
142
+ | 0.0727 | 1.16 | 46000 | 0.0651 | 0.6656 |
143
+ | 0.0768 | 1.17 | 46500 | 0.0641 | 0.6656 |
144
+ | 0.0722 | 1.18 | 47000 | 0.0640 | 0.6655 |
145
+ | 0.0763 | 1.19 | 47500 | 0.0646 | 0.6654 |
146
+ | 0.0766 | 1.21 | 48000 | 0.0636 | 0.6658 |
147
+ | 0.0774 | 1.22 | 48500 | 0.0636 | 0.6654 |
148
+ | 0.0759 | 1.23 | 49000 | 0.0633 | 0.6654 |
149
+ | 0.0779 | 1.24 | 49500 | 0.0625 | 0.6658 |
150
+ | 0.074 | 1.26 | 50000 | 0.0628 | 0.6654 |
151
+ | 0.0761 | 1.27 | 50500 | 0.0623 | 0.6656 |
152
+ | 0.0763 | 1.28 | 51000 | 0.0617 | 0.6655 |
153
+ | 0.072 | 1.29 | 51500 | 0.0617 | 0.6656 |
154
+ | 0.0718 | 1.31 | 52000 | 0.0618 | 0.6653 |
155
+ | 0.0703 | 1.32 | 52500 | 0.0611 | 0.6655 |
156
+ | 0.0718 | 1.33 | 53000 | 0.0608 | 0.6655 |
157
+ | 0.0686 | 1.34 | 53500 | 0.0610 | 0.6653 |
158
+ | 0.0688 | 1.36 | 54000 | 0.0604 | 0.6657 |
159
+ | 0.0694 | 1.37 | 54500 | 0.0604 | 0.6656 |
160
+ | 0.0736 | 1.38 | 55000 | 0.0598 | 0.6655 |
161
+ | 0.0674 | 1.39 | 55500 | 0.0599 | 0.6653 |
162
+ | 0.0681 | 1.41 | 56000 | 0.0592 | 0.6655 |
163
+ | 0.07 | 1.42 | 56500 | 0.0592 | 0.6653 |
164
+ | 0.0704 | 1.43 | 57000 | 0.0591 | 0.6656 |
165
+ | 0.0719 | 1.44 | 57500 | 0.0588 | 0.6653 |
166
+ | 0.0667 | 1.46 | 58000 | 0.0587 | 0.6653 |
167
+ | 0.0694 | 1.47 | 58500 | 0.0583 | 0.6653 |
168
+ | 0.0709 | 1.48 | 59000 | 0.0579 | 0.6655 |
169
+ | 0.0661 | 1.49 | 59500 | 0.0578 | 0.6655 |
170
+ | 0.0682 | 1.51 | 60000 | 0.0575 | 0.6655 |
171
+ | 0.0668 | 1.52 | 60500 | 0.0578 | 0.6654 |
172
+ | 0.0684 | 1.53 | 61000 | 0.0575 | 0.6653 |
173
+ | 0.0688 | 1.55 | 61500 | 0.0571 | 0.6652 |
174
+ | 0.068 | 1.56 | 62000 | 0.0572 | 0.6653 |
175
+ | 0.0694 | 1.57 | 62500 | 0.0566 | 0.6654 |
176
+ | 0.0642 | 1.58 | 63000 | 0.0569 | 0.6653 |
177
+ | 0.0646 | 1.6 | 63500 | 0.0564 | 0.6655 |
178
+ | 0.0633 | 1.61 | 64000 | 0.0566 | 0.6653 |
179
+ | 0.0677 | 1.62 | 64500 | 0.0563 | 0.6653 |
180
+ | 0.0649 | 1.63 | 65000 | 0.0560 | 0.6652 |
181
+ | 0.0654 | 1.65 | 65500 | 0.0558 | 0.6654 |
182
+ | 0.0675 | 1.66 | 66000 | 0.0557 | 0.6654 |
183
+ | 0.0642 | 1.67 | 66500 | 0.0554 | 0.6653 |
184
+ | 0.0631 | 1.68 | 67000 | 0.0552 | 0.6653 |
185
+ | 0.0628 | 1.7 | 67500 | 0.0552 | 0.6652 |
186
+ | 0.0658 | 1.71 | 68000 | 0.0550 | 0.6652 |
187
+ | 0.0654 | 1.72 | 68500 | 0.0547 | 0.6653 |
188
+ | 0.0648 | 1.73 | 69000 | 0.0544 | 0.6652 |
189
+ | 0.0634 | 1.75 | 69500 | 0.0547 | 0.6652 |
190
+ | 0.0642 | 1.76 | 70000 | 0.0544 | 0.6654 |
191
+ | 0.0649 | 1.77 | 70500 | 0.0542 | 0.6652 |
192
+ | 0.0641 | 1.78 | 71000 | 0.0540 | 0.6652 |
193
+ | 0.0659 | 1.8 | 71500 | 0.0540 | 0.6653 |
194
+ | 0.0651 | 1.81 | 72000 | 0.0536 | 0.6652 |
195
+ | 0.0625 | 1.82 | 72500 | 0.0536 | 0.6652 |
196
+ | 0.0631 | 1.83 | 73000 | 0.0536 | 0.6651 |
197
+ | 0.0614 | 1.85 | 73500 | 0.0535 | 0.6651 |
198
+ | 0.0637 | 1.86 | 74000 | 0.0533 | 0.6652 |
199
+ | 0.0619 | 1.87 | 74500 | 0.0532 | 0.6652 |
200
+ | 0.061 | 1.88 | 75000 | 0.0531 | 0.6652 |
201
+ | 0.0598 | 1.9 | 75500 | 0.0530 | 0.6652 |
202
+ | 0.0643 | 1.91 | 76000 | 0.0529 | 0.6652 |
203
+ | 0.0609 | 1.92 | 76500 | 0.0527 | 0.6651 |
204
+ | 0.06 | 1.93 | 77000 | 0.0527 | 0.6652 |
205
+ | 0.0627 | 1.95 | 77500 | 0.0527 | 0.6652 |
206
+ | 0.0607 | 1.96 | 78000 | 0.0526 | 0.6651 |
207
+ | 0.0607 | 1.97 | 78500 | 0.0525 | 0.6651 |
208
+ | 0.0608 | 1.98 | 79000 | 0.0525 | 0.6651 |
209
+ | 0.0609 | 2.0 | 79500 | 0.0525 | 0.6651 |
210
 
 
211
 
212
+ ### Framework versions
 
213
 
214
+ - Transformers 4.27.4
215
+ - Pytorch 2.0.0+cu117
216
+ - Datasets 2.11.0
217
+ - Tokenizers 0.13.2