File size: 1,989 Bytes
3d563af 6a40489 3d563af be873da 3d563af be873da 3d563af |
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 |
---
tags:
- generated_from_trainer
metrics:
- accuracy
base_model: Mathking/bert-base-german-cased-gnad10
model-index:
- name: bert-base-german-cased-gnad10-finetuned-tagesschau-subcategories
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. -->
# bert-base-german-cased-gnad10-finetuned-tagesschau-subcategories
This model is a fine-tuned version of [Mathking/bert-base-german-cased-gnad10](https://huggingface.co/Mathking/bert-base-german-cased-gnad10) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.4327
- Accuracy: 0.8733
## 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: 2e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
| 1.3737 | 0.4 | 30 | 0.9333 | 0.7133 |
| 0.7848 | 0.8 | 60 | 0.5591 | 0.8133 |
| 0.4933 | 1.2 | 90 | 0.4939 | 0.8 |
| 0.3441 | 1.6 | 120 | 0.5537 | 0.8133 |
| 0.3972 | 2.0 | 150 | 0.4229 | 0.8533 |
| 0.2103 | 2.4 | 180 | 0.4327 | 0.8733 |
| 0.1783 | 2.8 | 210 | 0.4834 | 0.8467 |
| 0.1367 | 3.2 | 240 | 0.4634 | 0.86 |
| 0.1273 | 3.6 | 270 | 0.4828 | 0.8467 |
### Framework versions
- Transformers 4.25.1
- Pytorch 1.12.1+cu113
- Datasets 2.7.1
- Tokenizers 0.13.2
|