update model card README.md

README.md

@@ -21,11 +21,11 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [bert-base-uncased](https://huggingface.co/bert-base-uncased) on the conll2003 dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.0714
-- Precision: 0.8968
-- Recall: 0.9138
-- F1: 0.9052
-- Accuracy: 0.9796
+- Loss: 0.0626
+- Precision: 0.9201
+- Recall: 0.9350
+- F1: 0.9275
+- Accuracy: 0.9832
 
 ## Model description
 
@@ -50,8 +50,8 @@ The following hyperparameters were used during training:
 - seed: 42
 - distributed_type: IPU
 - gradient_accumulation_steps: 16
-- total_train_batch_size: 64
-- total_eval_batch_size: 20
+- total_train_batch_size: 16
+- total_eval_batch_size: 5
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
 - num_epochs: 3
@@ -61,9 +61,9 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1     | Accuracy |
 |:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:|
-| 0.1048        | 1.0   | 219  | 0.0948          | 0.8606    | 0.8813 | 0.8708 | 0.9734   |
-| 0.1062        | 2.0   | 438  | 0.0742          | 0.8964    | 0.9072 | 0.9018 | 0.9789   |
-| 0.0458        | 3.0   | 657  | 0.0714          | 0.8968    | 0.9138 | 0.9052 | 0.9796   |
+| 0.0943        | 1.0   | 877  | 0.0687          | 0.9019    | 0.9149 | 0.9084 | 0.9801   |
+| 0.2395        | 2.0   | 1754 | 0.0623          | 0.9221    | 0.9298 | 0.9259 | 0.9829   |
+| 0.0241        | 3.0   | 2631 | 0.0626          | 0.9201    | 0.9350 | 0.9275 | 0.9832   |
 
 
 ### Framework versions