Update app.py

app.py

@@ -23,19 +23,17 @@ dtype = torch.float16
 
 processor = AutoProcessor.from_pretrained("StanfordAIMI/CheXagent-8b", trust_remote_code=True)
 generation_config = GenerationConfig.from_pretrained("StanfordAIMI/CheXagent-8b")
-model = AutoModelForCausalLM.from_pretrained("StanfordAIMI/CheXagent-8b", torch_dtype=dtype, trust_remote_code=True)
+# model = AutoModelForCausalLM.from_pretrained("StanfordAIMI/CheXagent-8b", torch_dtype=dtype, trust_remote_code=True)
 
 @spaces.GPU
 def generate(image, prompt):
+    model = AutoModelForCausalLM.from_pretrained("StanfordAIMI/CheXagent-8b", torch_dtype=dtype, trust_remote_code=True).to(device)
     if hasattr(image, "read"):
         image = Image.open(io.BytesIO(image.read())).convert("RGB")
     else:
         image = image
-
     images = [image]
-
     inputs = processor(images=images, text=f" USER: <s>{prompt} ASSISTANT: <s>", return_tensors="pt").to(device=device, dtype=dtype)
-
     output = model.generate(**inputs, generation_config=generation_config)[0]
     response = processor.tokenizer.decode(output, skip_special_tokens=True)
     return response