Create app.py

app.py

@@ -0,0 +1,53 @@
+import gradio as gr
+from transformers import AutoModelForCausalLM, AutoTokenizer
+import torch
+
+model_name = "microsoft/phi-3-mini-4k-instruct"
+
+tokenizer = AutoTokenizer.from_pretrained(model_name)
+model = AutoModelForCausalLM.from_pretrained(
+    model_name,
+    torch_dtype=torch.float16,
+    device_map="auto"
+)
+
+def chat_fn(message, history):
+    inputs = tokenizer.apply_chat_template(
+        history + [{"role": "user", "content": message}],
+        return_tensors="pt"
+    ).to(model.device)
+
+    outputs = model.generate(
+        inputs,
+        max_new_tokens=350,
+        temperature=0.7,
+        top_p=0.9
+    )
+
+    response = tokenizer.decode(outputs[0], skip_special_tokens=True)
+
+    # Clean to keep only assistant's last part
+    if "assistant" in response:
+        response = response.split("assistant")[-1].strip()
+
+    history.append({"role": "user", "content": message})
+    history.append({"role": "assistant", "content": response})
+
+    return response, history
+
+with gr.Blocks(title="Phi-3 Mini Demo") as demo:
+    gr.Markdown("## Phi-3 Mini 3.8B Instruct Demo")
+
+    with gr.Row():
+        chatbot = gr.Chatbot(height=420)
+
+    with gr.Row():
+        msg = gr.Textbox(label="Your message")
+
+    def user_send(message, history):
+        response, history = chat_fn(message, history)
+        return "", history
+
+    msg.submit(user_send, [msg, chatbot], [msg, chatbot])
+
+demo.launch()