Update README.md

README.md

@@ -44,11 +44,21 @@ The LLaMA 3.2 Multimodal News Media Bias Detector is a fine-tuned version of the
 
 ## How to Use
 
+**Sampled Data Usage**
+
+Please use the sampled data available at either of the following repositories for testing or development purposes with the LLaMA 3.2 Multimodal News Media Bias Detector:
+
+- [Sampled Data on Hugging Face - LLaMA 3.2 Multimodal News Media Bias Detector](https://huggingface.co/vector-institute/Llama3.2-Multimodal-Newsmedia-Bias-Detector/tree/main/sampled-data)
+- [NewsMediaBias-Plus Dataset on Hugging Face](https://huggingface.co/datasets/vector-institute/newsmediabias-plus)
+
+These datasets are pre-configured to be compatible with the model and include a variety of news articles and images that have been annotated for potential disinformation, allowing for effective model evaluation and demonstration.
+
+
 ### Installation
 
 Ensure you have the required libraries installed:
 
-```bash
+```
 pip install transformers accelerate bitsandbytes
 ```
 ```
@@ -249,7 +259,8 @@ Classification: Likely
 
 ## Training Data
 
-The model was fine-tuned on a custom dataset of news articles and images labeled for potential disinformation based on the presence of specific rhetorical techniques. The dataset includes balanced samples of articles classified as 'Likely' or 'Unlikely' to be disinformation.
+The model is fine-tuned on [NewsMediaBias-Plus Dataset on Hugging Face](https://huggingface.co/datasets/vector-institute/newsmediabias-plus)
+
 
 ## Training Procedure