Update app.py

app.py

@@ -97,6 +97,12 @@ model_x = Qwen2_5_VLForConditionalGeneration.from_pretrained(
     MODEL_ID_X, trust_remote_code=True, torch_dtype=torch.float16
 ).to(device).eval()
 
+MODEL_ID_Z = "Vchitect/ShotVL-3B"
+processor_z = AutoProcessor.from_pretrained(MODEL_ID_Z, trust_remote_code=True)
+model_z = Qwen2_5_VLForConditionalGeneration.from_pretrained(
+    MODEL_ID_Z, trust_remote_code=True, torch_dtype=torch.float16
+).to(device).eval()
+
 # --- PDF Generation and Preview Utility Function ---
 def generate_and_preview_pdf(image: Image.Image, text_content: str, font_size: int, line_spacing: float, alignment: str, image_size: str):
     """
@@ -189,6 +195,7 @@ def process_document_stream(
         
     if model_name == "LFM2-VL-450M": processor, model = processor_m, model_m
     elif model_name == "LFM2-VL-1.6B": processor, model = processor_t, model_t
+    elif model_name == "ShotVL-3B": processor, model = processor_z, model_z
     elif model_name == "SmolVLM-Instruct-250M": processor, model = processor_c, model_c
     elif model_name == "MonkeyOCR-pro-1.2B": processor, model = processor_g, model_g
     elif model_name == "VLAA-Thinker-Qwen2VL-2B": processor, model = processor_i, model_i
@@ -250,7 +257,7 @@ def create_gradio_interface():
             # Left Column (Inputs)
             with gr.Column(scale=1):
                 model_choice = gr.Dropdown(
-                    choices=["LFM2-VL-1.6B", "LFM2-VL-450M", "SmolVLM-Instruct-250M", "Megalodon-OCR-Sync-0713", "VLAA-Thinker-Qwen2VL-2B", "MonkeyOCR-pro-1.2B",
+                    choices=["LFM2-VL-1.6B", "LFM2-VL-450M", "SmolVLM-Instruct-250M", "ShotVL-3B", "Megalodon-OCR-Sync-0713", "VLAA-Thinker-Qwen2VL-2B", "MonkeyOCR-pro-1.2B",
                              "Nanonets-OCR-s"],
                     label="Select Model", value="LFM2-VL-1.6B"
                 )