Create app.py

app.py

@@ -0,0 +1,23 @@
+import gradio as gr
+from transformers import AutoModelForCausalLM, AutoTokenizer
+import torch
+
+model = AutoModelForCausalLM.from_pretrained("sakthi54321/Power_chat")
+tokenizer = AutoTokenizer.from_pretrained("sakthi54321/Power_chat")
+
+device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
+model = model.to(device)
+
+def generate(prompt):
+    inputs = tokenizer(f"Instruction: {prompt}\nOutput:", return_tensors="pt").to(device)
+    outputs = model.generate(**inputs, max_new_tokens=200)
+    response = tokenizer.decode(outputs[0], skip_special_tokens=True)
+    return response.split("Output:")[-1].strip()
+
+gr.Interface(
+    fn=generate,
+    inputs=gr.Textbox(lines=2, placeholder="Ask something..."),
+    outputs="text",
+    title="TinyLLaMA Chat",
+    description="Fine-tuned TinyLLaMA model on Alpaca/CodeAlpaca instructions."
+).launch()