Update app.py

app.py

@@ -1,84 +1,93 @@
 import gradio as gr
 import torch
 from PIL import Image
-# 根据你的模型，这里可能需要引入 transformers 中的 AutoModel, AutoProcessor 等
 # from transformers import AutoModel, AutoProcessor 
 
 # ==========================================
-# 1. 加载模型权重 (全局加载，避免每次点击重复加载)
+# 1. Load Model Weights 
 # ==========================================
 MODEL_ID = "yangxiaoda/SpatialLogic-ckpt-tag"
 device = "cuda" if torch.cuda.is_available() else "cpu"
 
-print(f"正在加载模型至 {device}...")
-# 【注意】这里需要替换为你模型真实的加载代码
-# 因为我无法看到该仓库内部的具体结构，通常代码如下：
+print(f"Loading model to {device}...")
+# [REPLACE THIS] with your actual model loading code
 # processor = AutoProcessor.from_pretrained(MODEL_ID)
 # model = AutoModel.from_pretrained(MODEL_ID).to(device)
-print("模型加载完成！")
+print("Model loaded successfully!")
 
 
 # ==========================================
-# 2. 定义核心推理函数
+# 2. Define Core Inference Function
 # ==========================================
 def compare_images(img1, img2, action_desc):
     """
-    img1: PIL.Image 格式
-    img2: PIL.Image 格式
-    action_desc: 字符串格式的动作描述
+    img1: PIL.Image
+    img2: PIL.Image
+    action_desc: string
     """
+    # 1. Input Validation
     if img1 is None or img2 is None:
-        return "请确保上传了两张图片！"
+        return "⚠️ Please ensure both images (img1 and img2) are uploaded!"
     if not action_desc.strip():
-        return "请输入动作描述！"
+        return "⚠️ Please enter an action description!"
 
     try:
-        # 【注意】这里需要替换为你模型真实的推理代码
-        # 伪代码示例：
+        # [REPLACE THIS] with your actual inference code
+        # Example pseudo-code:
         # inputs = processor(text=action_desc, images=[img1, img2], return_tensors="pt").to(device)
         # outputs = model(**inputs)
-        # score1, score2 = outputs.logits # 获取模型对两张图的打分
+        # score1, score2 = outputs.logits 
         
-        # 假设这里是你算出的得分（为了演示，这里用随机逻辑代替）
-        score_img1 = 0.8  # 替换为真实的 img1 得分
-        score_img2 = 0.4  # 替换为真实的 img2 得分
+        # Dummy scores for UI testing (Replace with real logic)
+        score_img1 = 0.8  
+        score_img2 = 0.4  
 
-        # 判断逻辑
+        # 2. Logic Judgment
         if score_img1 > score_img2:
             result = "img1"
-            explanation = "模型认为 **图片 1 (img1)** 更符合该动作描述。"
+            explanation = "The model evaluates that **Image 1 (img1)** is closer to the completed state of the described task."
         elif score_img2 > score_img1:
             result = "img2"
-            explanation = "模型认为 **图片 2 (img2)** 更符合该动作描述。"
+            explanation = "The model evaluates that **Image 2 (img2)** is closer to the completed state of the described task."
         else:
-            result = "平局"
-            explanation = "模型认为两张图完成度一致。"
+            result = "Tie"
+            explanation = "The model evaluates both images equally regarding the task completion."
 
-        return f"**输出结果:** {result}\n\n**详细说明:** {explanation}"
+        # 3. Format Output
+        return f"### **Output:** {result}\n\n**Explanation:** {explanation}"
     
     except Exception as e:
-        return f"推理过程中出错: {str(e)}"
+        return f"❌ An error occurred during inference: {str(e)}"
 
 # ==========================================
-# 3. 构建 Gradio 交互界面
+# 3. Build Gradio Interface
 # ==========================================
 with gr.Blocks(theme=gr.themes.Soft()) as demo:
-    gr.Markdown("# 🧠 SpatialLogic 空间逻辑推理演示")
-    gr.Markdown("上传两张图片并输入动作描述，模型将判断哪一张图片更接近任务完成状态。")
+    # Changed Title to the exact text you requested
+    gr.Markdown("# 🧠 SpatialLogic Reasoning Demo")
+    # Changed sub-title to English
+    gr.Markdown("Upload two images and input an action description. The model will predict which image is closer to the completed state of the task.")
     
     with gr.Row():
         with gr.Column():
-            img1_input = gr.Image(type="pil", label="图片 1 (img1)")
+            # Changed labels to English
+            img1_input = gr.Image(type="pil", label="Image 1 (img1)")
         with gr.Column():
-            img2_input = gr.Image(type="pil", label="图片 2 (img2)")
+            img2_input = gr.Image(type="pil", label="Image 2 (img2)")
             
-    text_input = gr.Textbox(label="动作描述 (Action Description)", placeholder="例如：把苹果放在桌子上")
+    # Changed label and placeholder to English
+    text_input = gr.Textbox(
+        label="Action Description", 
+        placeholder="e.g., Put the apple on the table"
+    )
     
-    submit_btn = gr.Button("🤖 开始判断", variant="primary")
+    # Changed button text to English
+    submit_btn = gr.Button("🤖 Evaluate", variant="primary")
     
-    output_text = gr.Markdown(label="判断结果")
+    # Output area
+    output_text = gr.Markdown(label="Result")
 
-    # 绑定函数与组件
+    # Bind function to components
     submit_btn.click(
         fn=compare_images, 
         inputs=[img1_input, img2_input, text_input], 
@@ -86,5 +95,4 @@ with gr.Blocks(theme=gr.themes.Soft()) as demo:
     )
 
 if __name__ == "__main__":
-    # 启动应用
     demo.launch()