Update app.py

app.py

@@ -3,8 +3,10 @@ import torch
 import gradio as gr
 
 model_id = "microsoft/phi-2"
+
+# Load tokenizer and model on CPU
 tokenizer = AutoTokenizer.from_pretrained(model_id)
-model = AutoModelForCausalLM.from_pretrained(model_id, torch_dtype=torch.float16, device_map="auto")
+model = AutoModelForCausalLM.from_pretrained(model_id)
 
 def chat(message):
     prompt = f"""
@@ -18,9 +20,11 @@ Avoid saying "as an AI" or sounding fake. Be real. Be humanlike. Be 𝕴 𝖆
 
 Now respond naturally to this message: {message}
     """
-    inputs = tokenizer(prompt, return_tensors="pt").to("cuda")
+    inputs = tokenizer(prompt, return_tensors="pt")
+    inputs = {k: v.to("cpu") for k, v in inputs.items()}  # make sure inputs are on CPU
     outputs = model.generate(**inputs, max_new_tokens=200)
     result = tokenizer.decode(outputs[0], skip_special_tokens=True)
+    # Return only the AI response part, removing prompt
     return result.split("Now respond naturally to this message:")[-1].strip()
 
 iface = gr.Interface(fn=chat, inputs="text", outputs="text")