Create app.py

app.py

@@ -0,0 +1,55 @@
+import gradio as gr
+from transformers import AutoTokenizer, AutoModelForSequenceClassification
+import torch
+import torch.nn.functional as F
+
+# Load model
+model_name = "tmt3103/BioASQ-yesno-PudMedBERT"
+tokenizer = AutoTokenizer.from_pretrained(model_name)
+model = AutoModelForSequenceClassification.from_pretrained(model_name)
+
+def predict_yesno(context, question):
+    inputs = tokenizer.encode_plus(
+        question,
+        context,
+        return_tensors="pt",
+        truncation=True,
+        max_length=512,
+        padding="max_length"
+    )
+
+    with torch.no_grad():
+        outputs = model(**inputs)
+        logits = outputs.logits
+        probs = F.softmax(logits, dim=1).squeeze()
+        pred_id = logits.argmax().item()
+
+    label = "Yes" if pred_id == 1 else "No"
+    return f"{label} (Confidence: {probs[pred_id]:.2f})"
+
+# Gradio
+with gr.Blocks() as demo:
+    gr.Markdown("#BioASQ Yes/No Question Answering")
+    gr.Markdown("""
+    This demo uses a fine-tuned BERT model to answer biomedical yes/no questions based on context.<br>
+    **Instructions**:
+    1. Paste the context (e.g., PubMed abstract).
+    2. Type your yes/no question.
+    3. Click 'Predict' to get the answer.
+    """)
+
+    with gr.Row():
+        with gr.Column():
+            context_input = gr.Textbox(label="Context", lines=8, placeholder="Paste biomedical context here...")
+            question_input = gr.Textbox(label="Yes/No Question", lines=2, placeholder="Enter your question here...")
+            predict_button = gr.Button("Predict")
+        with gr.Column():
+            output = gr.Textbox(label="Prediction")
+
+    predict_button.click(
+        fn=predict_yesno,
+        inputs=[context_input, question_input],
+        outputs=output
+    )
+
+demo.launch()