Update app.py

app.py

@@ -1,149 +1,203 @@
 import gradio as gr
 import torch
-from transformers import AutoTokenizer, AutoModelForCausalLM
+from transformers import AutoModel, AutoTokenizer
 import os
 import tempfile
 from PIL import Image, ImageDraw
 import re
 
-# -----------------------------------------
-# Force CPU Mode
-# -----------------------------------------
-device = torch.device("cpu")
-torch.set_default_device(device)
-
-# -----------------------------------------
-# 1. Load model ONCE at startup (CPU)
-# -----------------------------------------
-print("🔄 Loading model and tokenizer...")
+# --- 1. Load Model and Tokenizer (CPU only) ---
+print("Loading model and tokenizer on CPU...")
 model_name = "deepseek-ai/DeepSeek-OCR"
-
 tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True)
 
-model = AutoModelForCausalLM.from_pretrained(
+# Load model directly to CPU without flash_attention_2 (GPU-only feature)
+model = AutoModel.from_pretrained(
     model_name,
     trust_remote_code=True,
-    torch_dtype=torch.float32,
-    device_map={"": "cpu"}
+    use_safetensors=True,
+    torch_dtype=torch.float32  # Use float32 for CPU
 )
+model = model.eval()
+print("✅ Model loaded successfully on CPU.")
 
-model.eval()
-print("✅ Model loaded successfully (CPU mode)!")
-
-# -----------------------------------------
-# Helper: find generated result images
-# -----------------------------------------
+# --- Helper function to find pre-generated result images ---
 def find_result_image(path):
     for filename in os.listdir(path):
         if "grounding" in filename or "result" in filename:
             try:
-                return Image.open(os.path.join(path, filename))
-            except:
-                continue
+                image_path = os.path.join(path, filename)
+                return Image.open(image_path)
+            except Exception as e:
+                print(f"Error opening result image {filename}: {e}")
     return None
 
-# -----------------------------------------
-# 2. OCR main function
-# -----------------------------------------
+# --- 2. Main Processing Function (CPU version) ---
 def process_ocr_task(image, model_size, task_type, ref_text):
-
+    """
+    Processes an image with DeepSeek-OCR for all supported tasks.
+    CPU-only version without GPU decorators.
+    """
     if image is None:
-        return "Please upload image first.", None
-
-    print("⚙️ Running OCR (CPU mode)...")
-
-    # Prompt logic
-    if task_type == "📝 Free OCR":
-        prompt = "<image>\nFree OCR."
-    elif task_type == "📄 Convert to Markdown":
-        prompt = "<image>\n<|grounding|>Convert document to markdown."
-    elif task_type == "📈 Parse Figure":
-        prompt = "<image>\nParse the figure."
-    elif task_type == "🔍 Locate Object by Reference":
-        if not ref_text.strip():
-            raise gr.Error("Reference text required!")
-        prompt = f"<image>\nLocate <|ref|>{ref_text.strip()}<|/ref|> in the image."
-    else:
-        prompt = "<image>\nFree OCR."
-
-    # Size configs
-    size_configs = {
-        "Tiny": {"base_size": 512, "image_size": 512, "crop_mode": False},
-        "Small": {"base_size": 640, "image_size": 640, "crop_mode": False},
-        "Base": {"base_size": 1024, "image_size": 1024, "crop_mode": False},
-        "Large": {"base_size": 1280, "image_size": 1280, "crop_mode": False},
-        "Gundam (Recommended)": {"base_size": 1024, "image_size": 640, "crop_mode": True},
-    }
-    config = size_configs[model_size]
-
-    # Temporary image save
-    with tempfile.TemporaryDirectory() as output_path:
-        img_path = os.path.join(output_path, "input.png")
-        image.save(img_path)
+        return "Please upload an image first.", None
+
+    print("🚀 Processing on CPU...")
 
-        # ✅ Corrected infer API for DeepSeek-OCR
+    with tempfile.TemporaryDirectory() as output_path:
+        # Build the prompt
+        if task_type == "📝 Free OCR":
+            prompt = "<image>\nFree OCR."
+        elif task_type == "📄 Convert to Markdown":
+            prompt = "<image>\n<|grounding|>Convert the document to markdown."
+        elif task_type == "📈 Parse Figure":
+            prompt = "<image>\nParse the figure."
+        elif task_type == "🔍 Locate Object by Reference":
+            if not ref_text or ref_text.strip() == "":
+                raise gr.Error("For the 'Locate' task, you must provide the reference text to find!")
+            prompt = f"<image>\nLocate <|ref|>{ref_text.strip()}<|/ref|> in the image."
+        else:
+            prompt = "<image>\nFree OCR."
+
+        temp_image_path = os.path.join(output_path, "temp_image.png")
+        image.save(temp_image_path)
+
+        # Configure model size
+        size_configs = {
+            "Tiny": {"base_size": 512, "image_size": 512, "crop_mode": False},
+            "Small": {"base_size": 640, "image_size": 640, "crop_mode": False},
+            "Base": {"base_size": 1024, "image_size": 1024, "crop_mode": False},
+            "Large": {"base_size": 1280, "image_size": 1280, "crop_mode": False},
+            "Gundam (Recommended)": {"base_size": 1024, "image_size": 640, "crop_mode": True},
+        }
+        config = size_configs.get(model_size, size_configs["Gundam (Recommended)"])
+
+        print(f"🏃 Running inference with prompt: {prompt}")
+        
+        # Run inference on CPU (model is already on CPU)
         text_result = model.infer(
-            img=image,
+            tokenizer,
             prompt=prompt,
-            output_dir=output_path,
+            image_file=temp_image_path,
+            output_path=output_path,
             base_size=config["base_size"],
             image_size=config["image_size"],
-            crop_mode=config["crop_mode"]
+            crop_mode=config["crop_mode"],
+            save_results=True,
+            test_compress=True,
+            eval_mode=True,
         )
 
-        print("📜 Output text:", text_result[:200])
+        print(f"====\n📄 Text Result: {text_result}\n====")
 
-        # Detect bounding boxes if exist
+        # Try to find and draw all bounding boxes
+        result_image_pil = None
+        
+        # Pattern to find coordinates like [[280, 15, 696, 997]]
         pattern = re.compile(r"<\|det\|>\[\[(\d+),\s*(\d+),\s*(\d+),\s*(\d+)\]\]<\|/det\|>")
         matches = list(pattern.finditer(text_result))
 
         if matches:
-            result_img = image.copy()
-            draw = ImageDraw.Draw(result_img)
+            print(f"✅ Found {len(matches)} bounding box(es). Drawing on the original image.")
+            
+            # Create a copy of the original image to draw on
+            image_with_bboxes = image.copy()
+            draw = ImageDraw.Draw(image_with_bboxes)
             w, h = image.size
 
-            for m in matches:
-                x1n, y1n, x2n, y2n = map(int, m.groups())
-                draw.rectangle([
-                    int(x1n / 1000 * w),
-                    int(y1n / 1000 * h),
-                    int(x2n / 1000 * w),
-                    int(y2n / 1000 * h),
-                ], outline="red", width=3)
-
-            return text_result, result_img
-
-        return text_result, find_result_image(output_path)
-
-# -----------------------------------------
-# 3. UI Layout
-# -----------------------------------------
-with gr.Blocks(title="🐳DeepSeek-OCR🐳", theme=gr.themes.Soft()) as demo:
-    gr.Markdown("## DeepSeek-OCR Demo - CPU Mode")
+            for match in matches:
+                # Extract coordinates as integers
+                coords_norm = [int(c) for c in match.groups()]
+                x1_norm, y1_norm, x2_norm, y2_norm = coords_norm
+                
+                # Scale normalized coordinates to actual image size
+                x1 = int(x1_norm / 1000 * w)
+                y1 = int(y1_norm / 1000 * h)
+                x2 = int(x2_norm / 1000 * w)
+                y2 = int(y2_norm / 1000 * h)
+                
+                # Draw rectangle with red outline
+                draw.rectangle([x1, y1, x2, y2], outline="red", width=3)
+            
+            result_image_pil = image_with_bboxes
+        else:
+            print("⚠️ No bounding box coordinates found. Falling back to search for result image file.")
+            result_image_pil = find_result_image(output_path)
+            
+        return text_result, result_image_pil
+
+
+# --- 3. Build the Gradio Interface ---
+with gr.Blocks(title="🐳DeepSeek-OCR (CPU)🐳", theme=gr.themes.Soft()) as demo:
+    gr.Markdown(
+        """
+        # 🐳 DeepSeek-OCR (CPU Version) 🐳
+        **⚠️ Note: Running on CPU - processing will be slower than GPU version**
+        
+        **💡 How to use:**
+        1.  **Upload an image** using the upload box.
+        2.  Select a **Resolution**. Start with `Tiny` or `Small` for faster CPU processing.
+        3.  Choose a **Task Type**:
+            - **📝 Free OCR**: Extracts raw text from the image.
+            - **📄 Convert to Markdown**: Converts the document into Markdown.
+            - **📈 Parse Figure**: Extracts structured data from charts.
+            - **🔍 Locate Object by Reference**: Finds a specific object/text.
+        4. If this helpful, please give it a like! 🙏 ❤️
+        """
+    )
 
     with gr.Row():
         with gr.Column(scale=1):
-            image_input = gr.Image(type="pil", label="Upload Image")
+            image_input = gr.Image(type="pil", label="🖼️ Upload Image", sources=["upload", "clipboard"])
             model_size = gr.Dropdown(
-                ["Tiny", "Small", "Base", "Large", "Gundam (Recommended)"],
-                value="Gundam (Recommended)"
+                choices=["Tiny", "Small", "Base", "Large", "Gundam (Recommended)"], 
+                value="Small",  # Default to Small for faster CPU processing
+                label="⚙️ Resolution Size"
             )
             task_type = gr.Dropdown(
-                ["📝 Free OCR", "📄 Convert to Markdown", "📈 Parse Figure", "🔍 Locate Object by Reference"],
-                value="📄 Convert to Markdown"
+                choices=["📝 Free OCR", "📄 Convert to Markdown", "📈 Parse Figure", "🔍 Locate Object by Reference"], 
+                value="📄 Convert to Markdown", 
+                label="🚀 Task Type"
+            )
+            ref_text_input = gr.Textbox(
+                label="📝 Reference Text (for Locate task)", 
+                placeholder="e.g., the teacher, 20-10, a red car...", 
+                visible=False
             )
-            ref_text = gr.Textbox(visible=False)
-            btn = gr.Button("🚀 Process")
+            submit_btn = gr.Button("Process Image", variant="primary")
 
         with gr.Column(scale=2):
-            out_text = gr.Textbox(lines=12, show_copy_button=True)
-            out_image = gr.Image(type="pil", label="Result")
-
-    def toggle(t):
-        return gr.Textbox(visible=(t == "🔍 Locate Object by Reference"))
-
-    task_type.change(toggle, task_type, ref_text)
-    btn.click(process_ocr_task, [image_input, model_size, task_type, ref_text], [out_text, out_image])
-
+            output_text = gr.Textbox(label="📄 Text Result", lines=15, show_copy_button=True)
+            output_image = gr.Image(label="🖼️ Image Result (if any)", type="pil")
+
+    # UI Interaction Logic
+    def toggle_ref_text_visibility(task):
+        return gr.Textbox(visible=True) if task == "🔍 Locate Object by Reference" else gr.Textbox(visible=False)
+
+    task_type.change(fn=toggle_ref_text_visibility, inputs=task_type, outputs=ref_text_input)
+    submit_btn.click(
+        fn=process_ocr_task, 
+        inputs=[image_input, model_size, task_type, ref_text_input], 
+        outputs=[output_text, output_image]
+    )
+
+    # Examples
+    gr.Examples(
+        examples=[
+            ["doc_markdown.png", "Small", "📄 Convert to Markdown", ""],
+            ["chart.png", "Small", "📈 Parse Figure", ""],
+            ["teacher.jpg", "Tiny", "🔍 Locate Object by Reference", "the teacher"],
+            ["math_locate.jpg", "Tiny", "🔍 Locate Object by Reference", "20-10"],
+            ["receipt.jpg", "Small", "📝 Free OCR", ""],
+        ],
+        inputs=[image_input, model_size, task_type, ref_text_input],
+        outputs=[output_text, output_image],
+        fn=process_ocr_task,
+        cache_examples=False,
+    )
+
+# --- 4. Launch the App ---
 if __name__ == "__main__":
-    demo.queue().launch(server_name="0.0.0.0", server_port=7860)
+    if not os.path.exists("examples"):
+        os.makedirs("examples")
+    
+    demo.queue(max_size=5).launch(share=True)  # Reduced queue size for CPU