Update app.py

app.py

@@ -1,31 +1,101 @@
+import gradio as gr
+from huggingface_hub import InferenceClient
 import streamlit as st
 from transformers import pipeline
 from datasets import load_dataset
 
-# Initialize text-generation pipeline with the model
+# Initialize the Hugging Face InferenceClient
+client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
+
+# Initialize text-generation pipeline with the model for Streamlit
 model_name = "nvidia/Llama-3.1-Nemotron-70B-Instruct-HF"
 pipe = pipeline("text-generation", model=model_name)
 
 # Load the dataset
 ds = load_dataset("refugee-law-lab/canadian-legal-data", "default", split="train")
 
-# 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.")
 
-# Show dataset sample
-st.subheader("Sample Data from Canadian Legal Dataset:")
-st.write(ds[:5])  # Displaying the first 5 rows of the dataset
+# Gradio Function
+def respond(
+    message,
+    history: list[tuple[str, str]],
+    system_message,
+    max_tokens,
+    temperature,
+    top_p,
+):
+    messages = [{"role": "system", "content": system_message}]
+
+    for val in history:
+        if val[0]:
+            messages.append({"role": "user", "content": val[0]})
+        if val[1]:
+            messages.append({"role": "assistant", "content": val[1]})
+
+    messages.append({"role": "user", "content": message})
+
+    response = ""
+
+    for message in client.chat_completion(
+        messages,
+        max_tokens=max_tokens,
+        stream=True,
+        temperature=temperature,
+        top_p=top_p,
+    ):
+        token = message.choices[0].delta.content
+
+        response += token
+        yield response
+
+
+# Gradio interface setup
+demo = gr.ChatInterface(
+    respond,
+    additional_inputs=[
+        gr.Textbox(value="You are a friendly Chatbot.", 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)",
+        ),
+    ],
+)
+
+
+# Streamlit interface setup
+def 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.")
+
+    # Show dataset sample
+    st.subheader("Sample Data from Canadian Legal Dataset:")
+    st.write(ds[:5])  # Displaying the first 5 rows of the dataset
+
+    # 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.")
+
 
-# Prompt input
-prompt = st.text_area("Enter your prompt:", placeholder="Type something...")
+# Running Gradio and Streamlit interfaces
+if __name__ == "__main__":
+    st.sidebar.title("Choose an Interface")
+    interface = st.sidebar.radio("Select", ("Streamlit", "Gradio"))
 
-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)
+    if interface == "Streamlit":
+        streamlit_interface()
     else:
-        st.write("Please enter a prompt to generate a response.")
+        demo.launch()