Update app.py

app.py

@@ -9,8 +9,18 @@ from llava.constants import IMAGE_TOKEN_INDEX, DEFAULT_IMAGE_TOKEN
 from llava.model.builder import load_pretrained_model
 from llava.mm_utils import tokenizer_image_token
 from transformers.generation.streamers import TextIteratorStreamer
+import spaces
 
-# import spaces
+import subprocess
+
+if torch.cuda.is_available():
+    device = "cuda"
+    print("Using GPU")
+else:
+    device = "cpu"
+    print("Using CPU")
+
+subprocess.run(["pip", "install", "."])
 
 device = "cuda:0"
 tokenizer, model, image_processor, context_len = load_pretrained_model(model_path="./checkpoints/llava-phi3-3.8b-lora", model_name="llava-phi3-3.8b-lora", model_base="microsoft/Phi-3-mini-128k-instruct", load_8bit=False, load_4bit=False, device=device)
@@ -24,7 +34,7 @@ def load_image(image_file):
         image = Image.open(image_file).convert('RGB')
     return image
 
-# @spaces.GPU
+@spaces.GPU
 def bot_streaming(message, history):
     print(message)
     if message["files"]: