Update app.py

app.py

@@ -6,11 +6,11 @@ import gradio as gr
 # Load the model
 model = AutoModel.from_pretrained(
     'OpenGVLab/InternVL2_5-1B',
-    torch_dtype=torch.bfloat16,
+    torch_dtype=torch.float32,  # Use float32 for CPU compatibility
     low_cpu_mem_usage=True,
     trust_remote_code=True,
     use_flash_attn=False  # Disable Flash Attention
-).cuda().eval()
+).eval()  # Do not move to CUDA, leave on CPU
 
 # Load the image processor
 image_processor = CLIPImageProcessor.from_pretrained('OpenGVLab/InternVL2_5-1B')
@@ -23,9 +23,8 @@ def process_image(image):
         
         # Preprocess the image
         pixel_values = image_processor(images=image, return_tensors='pt').pixel_values
-        pixel_values = pixel_values.to(torch.bfloat16).cuda()
         
-        # Run the model
+        # Run the model on CPU
         outputs = model(pixel_values)
         
         # Assuming the model returns embeddings or features