Update app.py

app.py

@@ -1,22 +1,26 @@
 import gradio as gr
 from transformers import AutoTokenizer, AutoModelForSeq2SeqLM, pipeline
 
+# Load model from local files
 tokenizer = AutoTokenizer.from_pretrained(".", local_files_only=True)
 model = AutoModelForSeq2SeqLM.from_pretrained(".", local_files_only=True)
 pipe = pipeline("text2text-generation", model=model, tokenizer=tokenizer)
 
 def predict(task, prompt, context="", auto_cot=False):
-    input_str = f"[TASK: {task.upper()}] {prompt}"
+    full_prompt = f"[TASK: {task.upper()}] {prompt}"
     if context:
-        input_str += f" Context: {context}"
+        full_prompt += f" Context: {context}"
     if auto_cot:
-        input_str += "\nLet's think step by step."
-    return pipe(input_str, max_new_tokens=128)[0]["generated_text"]
+        full_prompt += "\nLet's think step by step."
 
+    output = pipe(full_prompt, max_new_tokens=128)[0]["generated_text"]
+    return output
+
+# ✅ API-only launch
 app = gr.Interface(
     fn=predict,
     inputs=["text", "text", "text", "checkbox"],
-    outputs="text"
+    outputs="text",
 )
 
-app.launch()
+app.launch(share=True, inline=False, show_api=True)