File size: 4,573 Bytes
fb907a4 |
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 |
---
library_name: transformers
license: mit
base_model: bert-base-german-cased
tags:
- generated_from_trainer
metrics:
- accuracy
- precision
- recall
- f1
model-index:
- name: classifier-de2
results: []
---
<!-- 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. -->
# classifier-de2
This model is a fine-tuned version of [bert-base-german-cased](https://huggingface.co/bert-base-german-cased) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.3294
- Accuracy: 0.8826
- Precision: 0.5399
- Recall: 0.3576
- F1: 0.4302
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 1.5e-05
- train_batch_size: 256
- eval_batch_size: 256
- seed: 42
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 3
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1 |
|:-------------:|:------:|:-----:|:---------------:|:--------:|:---------:|:------:|:------:|
| 0.2757 | 0.0923 | 900 | 0.3526 | 0.8732 | 0.4426 | 0.0879 | 0.1466 |
| 0.2537 | 0.1845 | 1800 | 0.3498 | 0.8739 | 0.4782 | 0.1823 | 0.2640 |
| 0.2242 | 0.2768 | 2700 | 0.3381 | 0.8815 | 0.5739 | 0.1712 | 0.2637 |
| 0.2061 | 0.3690 | 3600 | 0.3430 | 0.8763 | 0.5022 | 0.2519 | 0.3355 |
| 0.1914 | 0.4613 | 4500 | 0.3435 | 0.8784 | 0.5202 | 0.2482 | 0.3360 |
| 0.1798 | 0.5535 | 5400 | 0.3240 | 0.8817 | 0.5554 | 0.2291 | 0.3243 |
| 0.1899 | 0.6458 | 6300 | 0.3206 | 0.8768 | 0.5052 | 0.3153 | 0.3883 |
| 0.1761 | 0.7380 | 7200 | 0.3340 | 0.8846 | 0.5955 | 0.2170 | 0.3181 |
| 0.189 | 0.8303 | 8100 | 0.3241 | 0.8860 | 0.6141 | 0.2160 | 0.3196 |
| 0.1644 | 0.9225 | 9000 | 0.3278 | 0.8861 | 0.6105 | 0.2251 | 0.3289 |
| 0.1582 | 1.0148 | 9900 | 0.3437 | 0.8847 | 0.5773 | 0.2633 | 0.3616 |
| 0.1511 | 1.1070 | 10800 | 0.3187 | 0.8836 | 0.5556 | 0.3076 | 0.3960 |
| 0.1602 | 1.1993 | 11700 | 0.3198 | 0.8860 | 0.5858 | 0.2764 | 0.3756 |
| 0.149 | 1.2915 | 12600 | 0.3244 | 0.8842 | 0.5635 | 0.2945 | 0.3868 |
| 0.1512 | 1.3838 | 13500 | 0.3281 | 0.8863 | 0.5792 | 0.3040 | 0.3987 |
| 0.1463 | 1.4760 | 14400 | 0.3228 | 0.8869 | 0.5947 | 0.2753 | 0.3763 |
| 0.1372 | 1.5683 | 15300 | 0.3300 | 0.8872 | 0.5869 | 0.3048 | 0.4012 |
| 0.1545 | 1.6605 | 16200 | 0.3229 | 0.8866 | 0.5807 | 0.3086 | 0.4030 |
| 0.1755 | 1.7528 | 17100 | 0.3070 | 0.8854 | 0.5652 | 0.3280 | 0.4151 |
| 0.1403 | 1.8450 | 18000 | 0.3212 | 0.8877 | 0.5995 | 0.2836 | 0.3851 |
| 0.1425 | 1.9373 | 18900 | 0.3179 | 0.8861 | 0.5722 | 0.3235 | 0.4133 |
| 0.1271 | 2.0295 | 19800 | 0.3483 | 0.8843 | 0.5545 | 0.3411 | 0.4224 |
| 0.1235 | 2.1218 | 20700 | 0.3362 | 0.8858 | 0.5685 | 0.3294 | 0.4171 |
| 0.1324 | 2.2140 | 21600 | 0.3294 | 0.8826 | 0.5399 | 0.3576 | 0.4302 |
| 0.1236 | 2.3063 | 22500 | 0.3345 | 0.8859 | 0.5712 | 0.3214 | 0.4113 |
| 0.1264 | 2.3985 | 23400 | 0.3575 | 0.8876 | 0.5879 | 0.3141 | 0.4094 |
| 0.1157 | 2.4908 | 24300 | 0.3405 | 0.8872 | 0.5863 | 0.3058 | 0.4020 |
| 0.1261 | 2.5830 | 25200 | 0.3372 | 0.8874 | 0.5853 | 0.3165 | 0.4109 |
| 0.1346 | 2.6753 | 26100 | 0.3398 | 0.8863 | 0.5747 | 0.3205 | 0.4115 |
| 0.1099 | 2.7675 | 27000 | 0.3492 | 0.8872 | 0.5843 | 0.3122 | 0.4070 |
| 0.1295 | 2.8598 | 27900 | 0.3374 | 0.8871 | 0.5813 | 0.3191 | 0.4120 |
| 0.1259 | 2.9520 | 28800 | 0.3410 | 0.8875 | 0.5863 | 0.3152 | 0.4100 |
### Framework versions
- Transformers 4.51.3
- Pytorch 2.7.0+cu126
- Datasets 3.5.0
- Tokenizers 0.21.1
|