Update README.md

README.md

@@ -34,7 +34,10 @@ It is designed to support the development of open language models and for lingui
 
 ## Register labels and their abbreviations
 
-Below is a list of the register labels predicted by the model. Note that some labels are hierarchical; when a sublabel is predicted, its parent label is also predicted. For a more detailed description, see [here](https://turkunlp.org/register-annotation-docs/).
+Below is a list of the register labels predicted by the model. Note that some labels are hierarchical; when a sublabel is predicted, its parent label is also predicted. 
+For a more detailed description, see [here](https://turkunlp.org/register-annotation-docs/).
+
+The main labels are uppercase. To only include these main labels in the predictions, simply slice the model's output to keep only the uppercase labels.
 
 - **MT:** Machine translated or generated
 - **LY:** Lyrical
@@ -113,7 +116,7 @@ The model was trained using the Multilingual CORE Corpora, which will be publish
 #### Training Hyperparameters
 
 - **Batch size:** 8
-- **Epochs:** 7
+- **Epochs:** 21
 - **Learning rate:** 0.00005
 - **Precision:** bfloat16 (non-mixed precision)
 - **TF32:** Enabled
@@ -126,7 +129,19 @@ Average inference time (across 1000 iterations), using a single NVIDIA A100 GPU
 
 ## Evaluation
 
-Coming soon
+**Evaluation results (micro-F1 for the languages the models was trained on):**
+
+| Language | F1 (All labels) | F1 (Main labels) |
+| -------- | --------------- | ---------------- |
+| English  | 0.72            | 
+| Finnish  | 0.79            | 
+| French   | 0.75            | 
+| Swedish  | 0.81            | 
+| Turkish  | 0.77            | 
+ 
+**Zero-shot evaluation results (micro-F1):**
+
+
 
 
 ## Technical Specifications