Update README.md

README.md

@@ -6,6 +6,37 @@ license: mit
 
 > Merged and simplified dialog act datasets from the [silicone collection](https://huggingface.co/datasets/silicone/)
 
+All of the subsets of the original collection have been filtered (for errors and ambiguous classes), merged together and grouped into pairs of dialog turns. It is hypothesized that training dialog act classifier by including the previous utterance can help models pick up additional contextual cues and be better at inference esp if an utterance pair is provided.
+
+## Example training script
+
+```python
+from datasets import load_dataset
+from simpletransformers.classification import (
+    ClassificationModel, ClassificationArgs
+)
+
+# Get data
+silicone_merged = load_dataset("diwank/silicone-merged")
+
+train_df = silicone_merged["train"]
+eval_df = silicone_merged["validation"]
+
+model_args = ClassificationArgs(
+    num_train_epochs=8,
+    model_type="deberta", 
+    model_name="microsoft/deberta-large",
+    use_multiprocessing=False,
+    evaluate_during_training=True,
+)
+
+# Create a ClassificationModel
+model = ClassificationModel("deberta", "microsoft/deberta-large", args=model_args, num_labels=11)  # 11 labels in this dataset
+
+# Train model
+model.train_model(train_df, eval_df=eval_df)
+```
+
 **Note**: This dataset is highly imbalanced and it is recommended to use a library like [imbalanced-learn](https://imbalanced-learn.org/stable/) before proceeding with training.
 
 ## Feature description
@@ -34,6 +65,8 @@ license: mit
 
 *****
 
+## Appendix
+
 ### How the original datasets were mapped:
 
 ```python