Add instructions

src/app.py

@@ -106,7 +106,12 @@ gpu_specs = get_gpu_specs()
 model_list_info = cache_model_list()
 
 _, col, _ = st.columns([1,3,1])
-with col.expander("Information", expanded=True):
+with col.expander("Info", expanded=True):
+    st.markdown("""- Model can be selected from the drop-down at the right, press "enter" after selecting to refresh.
+- Results are broken down by precision, this can be seen in the graph and the chart below.""")
+
+_, col, _ = st.columns([1,3,1])
+with col.expander("Detailed Information", expanded=False):
     st.markdown("""- GPU information comes from [TechPowerUp GPU Specs](https://www.techpowerup.com/gpu-specs/)
 - Mainly based on [Model Memory Calculator by hf-accelerate](https://huggingface.co/spaces/hf-accelerate/model-memory-usage)
     using `transformers` library