Front end as FAQ test

app.py

@@ -302,6 +302,25 @@ def process_query(query, use_rag, stream=False):
 
 #--------------------------------------------------------------------------------------------------------------------------------
 
+FAQAllprompts = {
+    "Brainstorming for this space": [
+        ("My problem to solve is", "- please make 10 sub problems have to solve from this:"),
+    ],
+    "Expansive Problem solving": [
+        ("My problem to solve is", "- please make 10 sub problems have to solve from this:"),
+        ("My process to solve is", "- please make 10 sub processes have to solve from this:"),
+        ("My goal to solve is", "- please make 10 sub goals have to solve from this:"),
+        ("My task to solve is", "- please make 10 sub tasks have to solve from this:"),
+        ("My phase to solve is", "- please make 10 sub phases have to solve from this:"),
+        ("My component to solve is", "- please make 10 sub components have to solve from this:"),
+        ("My element to solve is", "- please make 10 sub elements have to solve from this:"),
+        ("A brief description of my current situation:", "- please list the most important task to pay attention to:"),
+        ("A brief description of my current situation to analyse:", "- please conduct a situational analysis:"),
+        ("A brief description of my current situation to decompose:", "- please conduct a problem decomposition:"),
+    ],
+}
+
+#--------------------------------------------------------------------------------------------------------------------------------
 # Default configuration template
 default_config = {
     'background': '/AutoGameBackgrounds/1stGameLoc123.png',
@@ -942,7 +961,19 @@ with gr.Blocks() as demo:
         <div style="width: 20%; text-align: center">HF + Gradio allows for api use so this my prototype tool for tool use test</div>
     </div>""")
     with gr.Accordion("Qwen 0.5B as Space Guide Tests", open=False):
-        gr.HTML("Placeholder for weak RAG Type - Explanations through an opensource embeddings engine")
+        with gr.Tab("General FAQ Attempt"):
+            FAQMainOutput = gr.TextArea(placeholder='Output will show here')
+            FAQCustomButtonInput = gr.TextArea(lines=1, placeholder='Prompt goes here')
+
+            for category_name, category_prompts in FAQAllprompts.items():
+                with gr.Accordion(f"General {category_name} Pattern based", open=False):
+                    with gr.Group():
+                        for index, (prompt, _) in enumerate(category_prompts):
+                            button = gr.Button(prompt)
+                            button.click(llmguide_generate_response, inputs=[FAQCustomButtonInput, gr.State(index), gr.State(category_name)], outputs=FAQMainOutput)
+
+        with gr.Tab("General RAG (Pathfinder?) Attempt"):
+            gr.HTML("Placeholder for weak RAG Type - Explanations through an opensource embeddings engine")
         # gr.Interface(
         #     fn=rag,
         #     inputs=[
@@ -958,46 +989,45 @@ with gr.Blocks() as demo:
         #     description="Ask a question and get an answer based on the retrieved context. The response is generated using a GPU-accelerated model. Resource usage is logged at the end of generation."
         # )
 
-        gr.Interface(
-            fn=process_query,
-            inputs=[
-                gr.Textbox(lines=2, placeholder="Enter your question here..."),
-                gr.Checkbox(label="Use RAG"),
-                gr.Checkbox(label="Stream output")
-            ],
-            outputs=[
-                gr.Textbox(label="Generated Response"),
-                gr.Textbox(label="Tokens per second"),
-                gr.Textbox(label="RAM Usage"),
-                gr.Textbox(label="Referenced Documents")
-            ],
-            title="RAG/Non-RAG Q&A System",
-            description="Ask a question with or without using RAG. The response is generated using a GPU-accelerated model. RAM usage and referenced document IDs (for RAG) are logged."
-        )
-        gr.HTML("Placeholder for FAQ type (merge as buttons on the above interface) - front end as prompt engineering for the first message to force direction of conversion")
-        
+            gr.Interface(
+                fn=process_query,
+                inputs=[
+                    gr.Textbox(lines=2, placeholder="Enter your question here..."),
+                    gr.Checkbox(label="Use RAG"),
+                    gr.Checkbox(label="Stream output")
+                ],
+                outputs=[
+                    gr.Textbox(label="Generated Response"),
+                    gr.Textbox(label="Tokens per second"),
+                    gr.Textbox(label="RAM Usage"),
+                    gr.Textbox(label="Referenced Documents")
+                ],
+                title="RAG/Non-RAG Q&A System",
+                description="Ask a question with or without using RAG. The response is generated using a GPU-accelerated model. RAM usage and referenced document IDs (for RAG) are logged."
+            )
 
-        gr.HTML("Placeholder for https://huggingface.co/h2oai/h2o-danube3-500m-chat-GGUF as alternative")
-        gr.HTML("Placeholder for qwen 2 72b as alternative use checkbox and gradio client api call")
-        gr.Markdown("# Qwen-0.5B-Instruct Language Model")
-        gr.Markdown("This demo uses the Qwen-0.5B-Instruct model to generate responses based on your input.")
-        gr.HTML("Example prompts: <br>I am writing a story about a chef. please write dishes to appear on the menu. <br>What are the most common decisions that a chef story would include? <br>What are the kinds problems that a chef story would include? <br>What are the kinds of out of reach goals that a chef story would include? <br>Continue this config - Paste any complete block of the config")
-        
-        with gr.Row():
-            with gr.Column():
-                llmguide_prompt = gr.Textbox(lines=2, placeholder="Enter your prompt here...")
-                llmguide_stream_checkbox = gr.Checkbox(label="Enable streaming")
-                llmguide_submit_button = gr.Button("Generate")
+        with gr.Tab("Any Request to Qwen2-0.5B"):
+            gr.HTML("Placeholder for https://huggingface.co/h2oai/h2o-danube3-500m-chat-GGUF as alternative")
+            gr.HTML("Placeholder for qwen 2 72b as alternative use checkbox and gradio client api call")
+            gr.Markdown("# Qwen-0.5B-Instruct Language Model")
+            gr.Markdown("This demo uses the Qwen-0.5B-Instruct model to generate responses based on your input.")
+            gr.HTML("Example prompts: <br>I am writing a story about a chef. please write dishes to appear on the menu. <br>What are the most common decisions that a chef story would include? <br>What are the kinds problems that a chef story would include? <br>What are the kinds of out of reach goals that a chef story would include? <br>Continue this config - Paste any complete block of the config")
             
-            with gr.Column():
-                llmguide_output = gr.Textbox(lines=10, label="Generated Response")
-                llmguide_tokens_per_second = gr.Textbox(label="Tokens per Second")
-        
-        llmguide_submit_button.click(
-            llmguide_generate_response,
-            inputs=[llmguide_prompt, llmguide_stream_checkbox],
-            outputs=[llmguide_output, llmguide_tokens_per_second],
-        )
+            with gr.Row():
+                with gr.Column():
+                    llmguide_prompt = gr.Textbox(lines=2, placeholder="Enter your prompt here...")
+                    llmguide_stream_checkbox = gr.Checkbox(label="Enable streaming")
+                    llmguide_submit_button = gr.Button("Generate")
+                
+                with gr.Column():
+                    llmguide_output = gr.Textbox(lines=10, label="Generated Response")
+                    llmguide_tokens_per_second = gr.Textbox(label="Tokens per Second")
+            
+            llmguide_submit_button.click(
+                llmguide_generate_response,
+                inputs=[llmguide_prompt, llmguide_stream_checkbox],
+                outputs=[llmguide_output, llmguide_tokens_per_second],
+            )
 
     Frontendpromptengforguide = """Suggest descriptions for media to fill the empty media fields -