<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
BACnet-Klassifizierung-Sanitaertechnik-bert-base-german-cased
This model is a fine-tuned version of bert-base-german-cased on the gart-labor "klassifizierung_sanitaer_v2" dataset. It achieves the following results on the evaluation set:
- Loss: 0.0039
- F1: [1. 1. 1.]
Model description
This model makes it possible to classify the sanitary technology components described with the BACnet standard into different categories.
The model is based on a German-language data set.
Intended uses & limitations
The model divides descriptive texts into the following sanitary engineering categories: Other, pressure boosting system, softening system, lifting system, sanitary_general, waste water, drinking water heating system and water meter.
Training and evaluation data
The model is based on a German-language data set.
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 16
- total_train_batch_size: 128
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 40.0
Training results
Training Loss | Epoch | Step | Validation Loss | F1 |
---|---|---|---|---|
0.0507 | 1.0 | 1 | 0.1080 | [1. 1. 1.] |
0.0547 | 2.0 | 2 | 0.0589 | [1. 1. 1.] |
0.0407 | 3.0 | 3 | 0.0427 | [1. 1. 1.] |
0.0294 | 4.0 | 4 | 0.0465 | [1. 1. 1.] |
0.0284 | 5.0 | 5 | 0.0291 | [1. 1. 1.] |
0.0208 | 6.0 | 6 | 0.0232 | [1. 1. 1.] |
0.0171 | 7.0 | 7 | 0.0198 | [1. 1. 1.] |
0.0153 | 8.0 | 8 | 0.0170 | [1. 1. 1.] |
0.0134 | 9.0 | 9 | 0.0144 | [1. 1. 1.] |
0.0126 | 10.0 | 10 | 0.0124 | [1. 1. 1.] |
0.0108 | 11.0 | 11 | 0.0109 | [1. 1. 1.] |
0.0096 | 12.0 | 12 | 0.0098 | [1. 1. 1.] |
0.0084 | 13.0 | 13 | 0.0089 | [1. 1. 1.] |
0.0082 | 14.0 | 14 | 0.0083 | [1. 1. 1.] |
0.0071 | 15.0 | 15 | 0.0077 | [1. 1. 1.] |
0.0068 | 16.0 | 16 | 0.0073 | [1. 1. 1.] |
0.0064 | 17.0 | 17 | 0.0069 | [1. 1. 1.] |
0.0059 | 18.0 | 18 | 0.0065 | [1. 1. 1.] |
0.0053 | 19.0 | 19 | 0.0061 | [1. 1. 1.] |
0.0052 | 20.0 | 20 | 0.0058 | [1. 1. 1.] |
0.005 | 21.0 | 21 | 0.0056 | [1. 1. 1.] |
0.0047 | 22.0 | 22 | 0.0053 | [1. 1. 1.] |
0.0044 | 23.0 | 23 | 0.0051 | [1. 1. 1.] |
0.0042 | 24.0 | 24 | 0.0050 | [1. 1. 1.] |
0.0043 | 25.0 | 25 | 0.0048 | [1. 1. 1.] |
0.004 | 26.0 | 26 | 0.0047 | [1. 1. 1.] |
0.004 | 27.0 | 27 | 0.0045 | [1. 1. 1.] |
0.004 | 28.0 | 28 | 0.0044 | [1. 1. 1.] |
0.0037 | 29.0 | 29 | 0.0044 | [1. 1. 1.] |
0.0037 | 30.0 | 30 | 0.0043 | [1. 1. 1.] |
0.0037 | 31.0 | 31 | 0.0042 | [1. 1. 1.] |
0.0035 | 32.0 | 32 | 0.0042 | [1. 1. 1.] |
0.0036 | 33.0 | 33 | 0.0041 | [1. 1. 1.] |
0.0035 | 34.0 | 34 | 0.0041 | [1. 1. 1.] |
0.0037 | 35.0 | 35 | 0.0040 | [1. 1. 1.] |
0.0034 | 36.0 | 36 | 0.0040 | [1. 1. 1.] |
0.0033 | 37.0 | 37 | 0.0040 | [1. 1. 1.] |
0.0034 | 38.0 | 38 | 0.0040 | [1. 1. 1.] |
0.0034 | 39.0 | 39 | 0.0040 | [1. 1. 1.] |
0.0034 | 40.0 | 40 | 0.0039 | [1. 1. 1.] |
Framework versions
- Transformers 4.21.1
- Pytorch 1.12.0+cu113
- Datasets 2.4.0
- Tokenizers 0.12.1