Update app.py

app.py

@@ -18,7 +18,7 @@ from threading import Thread
 tk = os.environ.get("HF_TOKEN")
 #login(tk)
 ckpt = "meta-llama/Llama-3.2-11B-Vision-Instruct"
-model = MllamaForConditionalGeneration.from_pretrained(ckpt,torch_dtype=torch.bfloat16).to("cuda")
+model = MllamaForConditionalGeneration.from_pretrained(ckpt,torch_dtype=torch.bfloat16).to("cpu")
 processor = AutoProcessor.from_pretrained(ckpt)
 r = sr.Recognizer()
 
@@ -143,7 +143,7 @@ def response(state:AppState = AppState()):
             if images == []:
                 inputs = processor(text=texts, return_tensors="pt").to("cuda")
             else:
-                inputs = processor(text=texts, images=images, return_tensors="pt").to("cuda")
+                inputs = processor(text=texts, images=images, return_tensors="pt").to("cpu")
             streamer = TextIteratorStreamer(processor, skip_special_tokens=True, skip_prompt=True)
             generation_kwargs = dict(inputs, streamer=streamer, max_new_tokens=max_new_tokens)
             thread = Thread(target=model.generate, kwargs=generation_kwargs)