Update app.py

app.py

@@ -3,11 +3,31 @@ import gradio as gr
 import torch  
 from transformers import AutoTokenizer, AutoModelForCausalLM
 
-title = """# Minitron-8B-Base"""
+title = """# Minitron-8B-Base Story Generator"""
 description = """
+# Minitron
+
 Minitron is a family of small language models (SLMs) obtained by pruning [NVIDIA's](https://huggingface.co/nvidia) Nemotron-4 15B model. We prune model embedding size, attention heads, and MLP intermediate dimension, following which, we perform continued training with distillation to arrive at the final models.
+
+# Short Story Generator
+Welcome to the Short Story Generator! This application helps you create unique short stories based on your inputs.
+
+**Instructions:**
+1. **Main Character:** Describe the main character of your story. For example, "a brave knight" or "a curious cat".
+2. **Setting:** Describe the setting where your story takes place. For example, "in an enchanted forest" or "in a bustling city".
+3. **Plot Twist:** Add an interesting plot twist to make the story exciting. For example, "discovers a hidden treasure" or "finds a secret portal to another world".
+
+After filling in these details, click the "Submit" button, and a short story will be generated for you.
 """
 
+inputs = [
+    gr.inputs.Textbox(label="Main Character", placeholder="e.g. a brave knight"),
+    gr.inputs.Textbox(label="Setting", placeholder="e.g. in an enchanted forest"),
+    gr.inputs.Textbox(label="Plot Twist", placeholder="e.g. discovers a hidden treasure")
+]
+
+outputs = gr.outputs.Textbox(label="Generated Story")
+
 # Load the tokenizer and model
 model_path = "nvidia/Minitron-8B-Base"
 tokenizer = AutoTokenizer.from_pretrained(model_path)
@@ -31,19 +51,39 @@ def respond(message, history, system_message, max_tokens, temperature, top_p):
 
     output_text = tokenizer.decode(output_ids[0], skip_special_tokens=True)
       
-    return output_text  
+    return output_text 
+
+@spaces.GPU 
+def generate_story(character, setting, plot_twist):
+    """Define the function to generate the story."""
+    prompt = f"Write a short story with the following details:\nMain character: {character}\nSetting: {setting}\nPlot twist: {plot_twist}\n\nStory:"
+    input_ids = tokenizer.encode(prompt, return_tensors="pt").to(model.device)
+
+    output_ids = model.generate(input_ids, max_length=50, num_return_sequences=1)
+
+    output_text = tokenizer.decode(output_ids[0], skip_special_tokens=True)
+      
+    return output_text 
   
-demo = gr.ChatInterface(
-    title=gr.Markdown(title),
-    # description=gr.Markdown(description),
-    fn=respond,  
-    additional_inputs=[
-        gr.Textbox(value="You are Minitron an AI assistant created by Tonic-AI", label="System message"),   
-        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),  
-        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),  
-        gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p (nucleus sampling)")  
-    ],  
-)  
+#demo = gr.ChatInterface(
+#    title=gr.Markdown(title),
+#    description=gr.Markdown(description),
+#    fn=generate_story,  
+#    additional_inputs=[
+#        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),  
+#        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),  
+#        gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p (nucleus sampling)")  
+#    ],  
+#)  
+
+# Create the Gradio interface
+demo = gr.Interface(
+    fn=generate_story,
+    inputs=inputs,
+    outputs=outputs,
+    title="Short Story Generator",
+    description=description
+)
   
 if __name__ == "__main__":  
     demo.launch()