Create app.py

app.py

@@ -0,0 +1,74 @@
+from transformers import pipeline, AutoModelForCausalLM, AutoTokenizer
+import gradio as gr
+
+# Load pre-trained Hugging Face model for recommendation tasks
+model_name = "TinyLlama/TinyLlama-1.1B-Chat-v1.0"  # Specialized model for general tasks
+tokenizer = AutoTokenizer.from_pretrained(model_name)
+model = AutoModelForCausalLM.from_pretrained(model_name)
+
+# Increase the tokenizer's max length to handle larger inputs
+tokenizer.model_max_length = 2048  # Increase this to a higher number if needed
+text_generator = pipeline("text-generation", model=model, tokenizer=tokenizer)
+
+# Function to handle laptop recommendation tasks
+def laptop_recommendation(user_input, task):
+    """
+    Handles laptop recommendation tasks based on user preferences.
+
+    Parameters:
+    - user_input: str, the input text from the user with laptop preferences.
+    - task: str, the type of task (e.g., "Recommendation", "Compare", "Budget Recommendation").
+
+    Returns:
+    - str: The generated response.
+    """
+    # Construct prompts based on the selected task
+    if task == "Recommendation":
+        prompt = f"Recommend a laptop based on the following preferences:\n{user_input}\nRecommended Laptop:"
+    elif task == "Compare":
+        prompt = f"Compare two laptops based on the following specifications:\n{user_input}\nComparison:"
+    elif task == "Budget Recommendation":
+        prompt = f"Recommend the best laptop for the following budget:\n{user_input}\nRecommended Laptop for Budget:"
+    else:
+        return "Invalid task selected."
+
+    # Generate response using the model with an increased max length
+    response = text_generator(
+        prompt,
+        max_length=96,  # You can adjust this to fit the response size you want
+        num_return_sequences=1,
+        pad_token_id=tokenizer.eos_token_id,
+        temperature=0.7,
+        top_p=0.9
+    )[0]["generated_text"]
+
+    # Extract only the generated response beyond the prompt
+    return response[len(prompt):].strip()
+
+# Gradio Interface
+def gradio_interface(user_input, task):
+    """
+    Interface function for Gradio integration.
+    """
+    if not user_input.strip():
+        return "Please enter some input."
+    return laptop_recommendation(user_input, task)
+
+with gr.Blocks() as laptop_recommendation_ui:
+    gr.Markdown("# Laptop Recommendation Chatbot")
+    gr.Markdown("This chatbot helps with recommending laptops based on preferences, comparing laptops, and suggesting options based on budget.")
+
+    # User input components
+    user_input = gr.Textbox(lines=5, placeholder="Enter your laptop preferences here...", label="Your Input")
+    task = gr.Radio(["Recommendation", "Compare", "Budget Recommendation"], label="Select Task")
+    output = gr.Textbox(lines=10, label="Chatbot Response")
+
+    # Buttons
+    submit_button = gr.Button("Submit")
+    clear_button = gr.Button("Clear")
+
+    # Interaction
+    submit_button.click(gradio_interface, inputs=[user_input, task], outputs=output)
+    clear_button.click(lambda: ("", ""), None, [user_input, output])
+
+laptop_recommendation_ui.launch()