update model card README.md

README.md

@@ -19,11 +19,11 @@ should probably proofread and complete it, then remove this comment. -->
 
 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: 0.0367
-- Precision: 0.7687
-- Recall: 0.8429
-- F1: 0.8041
-- Accuracy: 0.9916
+- Loss: 0.0456
+- Precision: 0.7190
+- Recall: 0.85
+- F1: 0.7791
+- Accuracy: 0.9904
 
 ## Model description
 
@@ -43,21 +43,23 @@ More information needed
 
 The following hyperparameters were used during training:
 - learning_rate: 2e-05
-- train_batch_size: 16
+- train_batch_size: 32
 - eval_batch_size: 32
 - seed: 42
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
-- num_epochs: 3
+- num_epochs: 10
 
 ### Training results
 
 | Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1     | Accuracy |
 |:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:|
-| No log        | 0.62  | 50   | 0.0382          | 0.8020    | 0.8393 | 0.8202 | 0.9914   |
-| No log        | 1.25  | 100  | 0.0381          | 0.7421    | 0.8429 | 0.7893 | 0.9904   |
-| No log        | 1.88  | 150  | 0.0380          | 0.7429    | 0.8357 | 0.7866 | 0.9906   |
-| No log        | 2.5   | 200  | 0.0367          | 0.7687    | 0.8429 | 0.8041 | 0.9916   |
+| No log        | 0.8   | 32   | 0.0442          | 0.8066    | 0.7893 | 0.7978 | 0.9903   |
+| No log        | 1.6   | 64   | 0.0435          | 0.7337    | 0.8464 | 0.7861 | 0.9884   |
+| No log        | 2.4   | 96   | 0.0366          | 0.7702    | 0.85   | 0.8081 | 0.9909   |
+| No log        | 3.2   | 128  | 0.0389          | 0.7697    | 0.8357 | 0.8014 | 0.9907   |
+| No log        | 4.0   | 160  | 0.0377          | 0.7664    | 0.8321 | 0.7979 | 0.9911   |
+| No log        | 4.8   | 192  | 0.0456          | 0.7190    | 0.85   | 0.7791 | 0.9904   |
 
 
 ### Framework versions