Update app.py

app.py

@@ -156,19 +156,17 @@ Use the provided context to generate a result based on the following JSON format
 Instructions:
 1. Focus on the user's specific request and avoid irrelevant ingredients or approaches.
 2. Do not return anything other than the JSON.
-3. If the answer is unclear or the context does not fully address the prompt, return [].
+3. If the answer is unclear AND the context does not fully address the prompt, return [].
 4. Base the response on simple, healthy, and accessible ingredients and techniques.
 5. Rewrite the description in third person
-6. If context does not match {question} at all, return []
-7. Include the ingredient amounts and say them in the steps.
+6. Include the ingredient amounts and say them in the steps.
 
 When choosing a recipe from the context, FOLLOW these instructions:
-0. If context does not match {question} at all, return []
 1. The recipe should be makeable from scratch, using only proper ingredients and not other dishes or pre-made recipes
 2. If the recipes from the context makes sense but do not match {question}, generate an amazing, specific recipe for {question} 
    with precise steps and measurements. Take some inspiration from context if availab.e
 3. Following the above template.
-4. If context does not match {question} at all, return []
+4. If context does not match {question} in topic, return []
 
 
 Context: {context}
@@ -335,47 +333,6 @@ def display_prediction_graph(class_names, confidences):
     # Display the chart in Streamlit
     st.plotly_chart(fig)
     
-# def display_prediction_graph(class_names, confidences):
-#     #reversing them so graph displays highest predictions at the top
-#     confidences.reverse()
-#     class_names.reverse()
-
-#     #display as a graph
-#     norm = plt.Normalize(min(confidences), max(confidences))
-#     cmap = LinearSegmentedColormap.from_list("grey_orange", ["#808080", "#FFA500"]) #color map grey to orange
-    
-#     fig, ax = plt.subplots(figsize=(12, 6))
-#     bars = ax.barh(class_names, confidences, color=cmap(norm(confidences)))
-
-#     fig.patch.set_alpha(0)  # Transparent background
-#     ax.set_facecolor('none') 
-
-#     min_width = 0.07 * ax.get_xlim()[1]  # 7% of the x-axis range
-#     # Add labels inside the bars, aligned to the right
-#     for bar in bars:
-#         original_width = bar.get_width()
-#         width = original_width
-#         if width < min_width:
-#             width = min_width
-#         ax.text(width - 0.02, bar.get_y() + bar.get_height()/2, f'{original_width:.1f}%', 
-#                 va='center', ha='right', color='white', fontweight='bold', fontsize=16)
-
-#     ax.set_xticklabels([]) #remove x label
-
-#     # Wrapping labels
-#     max_label_width = 10
-#     labels = ax.get_yticklabels()
-#     wrapped_labels = [textwrap.fill(label.get_text(), width=max_label_width) for label in labels] # Wrap the labels if they exceed the max width
-#     ax.set_yticklabels(wrapped_labels, fontsize=16, color='white')
-    
-#     #no borders
-#     for spine in ax.spines.values():
-#         spine.set_visible(False)
-
-#     ax.set_title(class_names[-1], color='white', fontsize=24, fontweight='bold', ha='left', x=0.5)
-    
-#     st.pyplot(fig) # Display the plot
-    
 # #Streamlit
 
 #Left sidebar title