Update app.py

app.py

@@ -1,27 +1,31 @@
-import gradio as gr
+import streamlit as st
 from transformers import pipeline
 from datasets import load_dataset
 
-# Initialize text-generation pipeline
+# Initialize text-generation pipeline with the model
 model_name = "nvidia/Llama-3.1-Nemotron-70B-Instruct-HF"
 pipe = pipeline("text-generation", model=model_name)
 
-# Load dataset (fetching the dataset via the API endpoint you provided)
-dataset = load_dataset("refugee-law-lab/canadian-legal-data", split="train")
+# Load the dataset
+ds = load_dataset("refugee-law-lab/canadian-legal-data", "default", split="train")
 
-def generate_response(prompt):
-    # Pass the prompt to the model pipeline
-    generated_text = pipe(prompt, max_length=100, do_sample=True, temperature=0.7)[0]["generated_text"]
-    return generated_text
+# Streamlit interface
+st.title("Canadian Legal Text Generator")
+st.write("Enter a prompt related to Canadian legal data and generate text using Llama-3.1.")
 
-# Set up Gradio interface (optional if you want a web app)
-demo = gr.Interface(
-    fn=generate_response,
-    inputs=gr.Textbox(lines=2, placeholder="Enter your question or prompt here..."),
-    outputs="text",
-    title="Canadian Legal Text Generator"
-)
+# Show dataset sample
+st.subheader("Sample Data from Canadian Legal Dataset:")
+st.write(ds[:5])  # Displaying the first 5 rows of the dataset
 
-# Launch app
-if __name__ == "__main__":
-    demo.launch()
+# Prompt input
+prompt = st.text_area("Enter your prompt:", placeholder="Type something...")
+
+if st.button("Generate Response"):
+    if prompt:
+        # Generate text based on the prompt
+        with st.spinner("Generating response..."):
+            generated_text = pipe(prompt, max_length=100, do_sample=True, temperature=0.7)[0]["generated_text"]
+            st.write("**Generated Text:**")
+            st.write(generated_text)
+    else:
+        st.write("Please enter a prompt to generate a response.")