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: 0.0475
-- Precision: 0.7476
-- Recall: 0.8464
-- F1: 0.7940
-- Accuracy: 0.9908
+- Loss: 0.0381
+- Precision: 0.7815
+- Recall: 0.8429
+- F1: 0.8110
+- Accuracy: 0.9914
 
 ## Model description
 
@@ -43,21 +43,20 @@ More information needed
 
 The following hyperparameters were used during training:
 - learning_rate: 2e-05
-- train_batch_size: 8
-- eval_batch_size: 8
+- train_batch_size: 16
+- eval_batch_size: 32
 - seed: 42
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
-- num_epochs: 4
+- num_epochs: 3
 
 ### Training results
 
 | Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1     | Accuracy |
 |:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:|
-| No log        | 1.0   | 160  | 0.0497          | 0.6704    | 0.8643 | 0.7551 | 0.9870   |
-| No log        | 2.0   | 320  | 0.0370          | 0.7912    | 0.8393 | 0.8146 | 0.9924   |
-| No log        | 3.0   | 480  | 0.0443          | 0.7660    | 0.8536 | 0.8074 | 0.9910   |
-| 0.0585        | 4.0   | 640  | 0.0475          | 0.7476    | 0.8464 | 0.7940 | 0.9908   |
+| No log        | 1.0   | 80   | 0.0443          | 0.7139    | 0.8464 | 0.7745 | 0.9889   |
+| No log        | 2.0   | 160  | 0.0396          | 0.7692    | 0.8571 | 0.8108 | 0.9912   |
+| No log        | 3.0   | 240  | 0.0381          | 0.7815    | 0.8429 | 0.8110 | 0.9914   |
 
 
 ### Framework versions