last commits

app.py

@@ -1,32 +1,22 @@
-import streamlit as st
+import gradio as gr
 from transformers import pipeline
 
-# Set page config FIRST — must be the first Streamlit command
-st.set_page_config(page_title="BERT QA", layout="wide")
+# Load pipeline
+pipe = pipeline("sentiment-analysis")
+# pipe = pipeline("text-classification", model="ProsusAI/finbert")
 
-# Load QA pipeline with BERT model
-@st.cache_resource
-def load_qa_pipeline():
-    model_name = "deepset/bert-large-uncased-whole-word-masking-squad2"
-    return pipeline("question-answering", model=model_name, tokenizer=model_name)
+# Define function for prediction
+def analyze(text):
+    return pipe(text)
 
-nlp = load_qa_pipeline()
+# Build Gradio Interface
+demo = gr.Interface(
+    fn=analyze,
+    inputs=gr.Textbox(label="Enter your input:", lines=3, placeholder="Type text here..."),
+    outputs=gr.JSON(label="Result"),
+    title="Sentiment Analysis",
+    description="Enter text and get sentiment analysis using Hugging Face Transformers."
+)
 
-# Streamlit UI
-st.title("🤖 BERT-based Q&A using HuggingFace")
-
-context = st.text_area("📄 Enter Context", height=200, value="""
-Trainer: Jai Ganesh S;
-Session 1: Intro about Arenas & LLMs;
-Session 2: Hands-on on LLMs
-Session 3: LangChain - Intro
-""")
-
-question = st.text_input("❓ Ask a Question", value="What's the session 1?")
-
-if st.button("Get Answer"):
-    with st.spinner("Thinking..."):
-        QA_input = {"question": question, "context": context}
-        result = nlp(QA_input)
-        st.success(f"🧠 **Answer**: {result['answer']}")
-        st.caption(f"Score: {result['score']:.2f} | Start: {result['start']}, End: {result['end']}")
+# Launch
+demo.launch()