Update app.py

app.py

@@ -1,3 +1,4 @@
+
 import base64
 import os
 import urllib.parse
@@ -77,7 +78,7 @@ caption_gen_pipe = hf_pipeline("text-generation", model="Qwen/Qwen2.5-0.5B-Instr
 vqa_processor = BlipProcessor.from_pretrained("Salesforce/blip-vqa-base")
 vqa_model = BlipForQuestionAnswering.from_pretrained("Salesforce/blip-vqa-base").to(DEVICE)
 
-print("Loading segmentation and Stable Diffusion pipelines...")
+print("Loading clothing segmentation and Stable Diffusion pipelines...")
 seg_processor = SegformerImageProcessor.from_pretrained("mattmdjaga/segformer_b2_clothes")
 seg_model = AutoModelForSemanticSegmentation.from_pretrained("mattmdjaga/segformer_b2_clothes").to(DEVICE)
 inpaint_pipe = StableDiffusionInpaintPipeline.from_pretrained(
@@ -143,18 +144,28 @@ def pil_to_base64(img, max_size=280):
     return base64.b64encode(buf.getvalue()).decode("utf-8")
 
 # ---------------------------------------------------------------------------
-# EMBEDDING + FAISS SEARCH
+# EMBEDDING + FAISS SEARCH (תיקון חילוץ הוקטורים מתוך אובייקט ה-CLIP)
 # ---------------------------------------------------------------------------
 @torch.no_grad()
 def embed_query_image(pil_image):
     inputs = win_processor(images=pil_image, return_tensors="pt").to(DEVICE)
-    feats = win_model.get_image_features(**inputs)
+    outputs = win_model.get_image_features(**inputs)
+    # תיקון קריטי: חילוץ וקטור המאפיינים האמיתי מתוך ה-BaseModelOutput
+    if hasattr(outputs, "image_embeds"):
+        feats = outputs.image_embeds
+    else:
+        feats = outputs
     return feats.cpu().numpy().astype("float32")
 
 @torch.no_grad()
 def embed_query_text(sentence):
     inputs = win_processor(text=[sentence], return_tensors="pt", padding=True, truncation=True).to(DEVICE)
-    feats = win_model.get_text_features(**inputs)
+    outputs = win_model.get_text_features(**inputs)
+    # תיקון קריטי: חילוץ וקטור המאפיינים האמיתי מתוך ה-BaseModelOutput
+    if hasattr(outputs, "text_embeds"):
+        feats = outputs.text_embeds
+    else:
+        feats = outputs
     return feats.cpu().numpy().astype("float32")
 
 def build_feature_sentence(skin_tone, undertone, style_preference, gender=None, age_group=None):
@@ -192,7 +203,7 @@ def faiss_filtered_search(query_emb, top_k=3, exclude_idx=None, gender=None, age
     return np.array(kept_i), df.iloc[kept_i], np.array(kept_d)
 
 # ---------------------------------------------------------------------------
-# PATTERN LOGIC & COMPONENT BUILDERS (FIXED SYNTAXERRORS & MISSING FUNCTIONS)
+# PATTERN LOGIC & COMPONENT BUILDERS
 # ---------------------------------------------------------------------------
 def generate_stylist_caption(row):
     user_prompt = (
@@ -236,8 +247,6 @@ def generate_new_outfit_image(pil_image, row, target_gender=None):
     resized_protect_img = Image.fromarray(protect.astype(np.uint8) * 255).resize((head_w, head_h), resample=Image.NEAREST)
     
     canvas = Image.new("RGB", (canvas_w, canvas_h), color=(240, 238, 235))
-    paste_x = (canvas_w - head_w) // 2
-    paste_y = int(canvas_h * 0.03)
     canvas.paste(resized_face_crop, (paste_x, paste_y))
     
     mask_arr = np.full((canvas_h, canvas_w), 255, dtype=np.uint8)
@@ -252,7 +261,6 @@ def generate_new_outfit_image(pil_image, row, target_gender=None):
     generated = inpaint_pipe(prompt=prompt, image=canvas, mask_image=mask, num_inference_steps=25, height=canvas_h, width=canvas_w).images[0]
     return Image.composite(generated, canvas, mask), prompt
 
-# --- CRITICAL FIX: DEFINING THE MISSING HEADER AND PROFILE CARD BUILDER ---
 def build_style_card_html(row, caption):
     colors = [c.strip() for c in str(row["recommended_colors"]).split(",") if c.strip()][:4]
     swatches = "".join(
@@ -285,7 +293,6 @@ def to_shop_link(retailer, value, gender=None):
     return f"https://www.google.com/search?q={retailer}+{query_enc}"
 
 def build_outfit_component_cards_html(row):
-    """Generates the clean 6 core component cards layout mirroring LookMatch perfectly"""
     colors = [c.strip() for c in str(row["recommended_colors"]).split(",") if c.strip()] or ["neutral"]
     components = [
         ("TOP", row.get("outfit_top", "Top"), "zara", "search_query_zara"),
@@ -440,7 +447,8 @@ h1, h2, h3, p, span, label, input, select, textarea, button { color: #2C2A29 !im
 # ---------------------------------------------------------------------------
 # INTERFACE BUILD
 # ---------------------------------------------------------------------------
-with gr.Blocks(title="Personal Color Styling", css=CUSTOM_CSS, theme=gr.themes.Soft(primary_hue="amber")) as demo:
+# CRITICAL FIX: Moving custom CSS and theme injection parameters directly to the final launch function to comply with Gradio 6
+with gr.Blocks(title="Personal Color Styling") as demo:
     gr.HTML("""
     <div style="text-align:center; padding:24px 20px 10px;">
       <div style="font-size:11px; font-weight:700; color:#D2527F; letter-spacing:.18em; text-transform:uppercase; margin-bottom:8px;">Personal Color Styling</div>
@@ -471,8 +479,9 @@ with gr.Blocks(title="Personal Color Styling", css=CUSTOM_CSS, theme=gr.themes.S
             [style_card_a, new_img_a, answer_a, outfit_cards_a],
         )
 
+        # CRITICAL FIX: All expected positional inputs must map corresponding elements to prevent load_example mismatches
         gr.Examples(
-            examples=[[p, None, None, "What style would suit me best?"] for p in SAMPLE_PHOTOS],
+            examples=[[p, "woman", "adult", "What style would suit me best?"] for p in SAMPLE_PHOTOS],
             inputs=[photo_in, gender_a, age_a, question_a],
             outputs=[style_card_a, new_img_a, answer_a, outfit_cards_a],
             fn=recommend_from_photo,
@@ -509,4 +518,5 @@ with gr.Blocks(title="Personal Color Styling", css=CUSTOM_CSS, theme=gr.themes.S
         )
 
 if __name__ == "__main__":
-    demo.launch()
+    # CRITICAL FIX: Safe execution of Gradio 6 theme parameters
+    demo.launch(css=CUSTOM_CSS, theme=gr.themes.Soft(primary_hue="amber"))