chore: Refactor generate_instruction_response function and optimize Gradio demo description

app.py

@@ -35,6 +35,10 @@ terminators = [
 
 @spaces.GPU
 def generate_instruction_response():
+    prompt_info = (
+        f"Generating a user prompt from the LLM via the template `{extract_input}`\n\n"
+    )
+    yield prompt_info
     instruction = pipeline(
         extract_input,
         max_new_tokens=2048,
@@ -48,8 +52,9 @@ def generate_instruction_response():
         len(extract_input) :
     ].split("\n")[0]
 
-    first_rep = "## LLm generated instruction:\n\n" + sanitized_instruction
-    yield first_rep + "\n\n generating LLM response..."
+    first_step = "## LLM generated instruction:\n\n" + sanitized_instruction
+    first_step = prompt_info + first_step
+    yield first_step + "\n\n generating LLM response..."
 
     response_template = f"""<|begin_of_text|><|start_header_id|>user<|end_header_id|>\n\n{sanitized_instruction}<|eot_id|><|start_header_id|>assistant<|end_header_id|>\n\n"""
 
@@ -64,7 +69,7 @@ def generate_instruction_response():
 
     assistant_response = response[0]["generated_text"][len(response_template) :]
 
-    yield (first_rep + "\n\n## LLM Generated response:\n\n" + assistant_response)
+    yield (first_step + "\n\n## LLM Generated response:\n\n" + assistant_response)
 
 
 title = "Magpie demo"