Update app.py

app.py

@@ -1,3 +1,5 @@
+# app.py
+
 import torch
 import warnings
 import numpy as np
@@ -17,18 +19,29 @@ from model import (
 warnings.filterwarnings("ignore", category=FutureWarning)
 warnings.filterwarnings("ignore", category=UserWarning)
 
+
+# ------------------------------------------------------------
+# 1) CİHAZ AYARI ve MODELLERİN ÖN YÜKLENMESİ
+# ------------------------------------------------------------
 device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
 
+# a) CNN modellerini yükle (Vbai-DPA-2.3 repo’su)
 cnn_model_f = load_classification_model(device, model_type="f", num_classes=6)
 cnn_model_c = load_classification_model(device, model_type="c", num_classes=6)
 cnn_model_q = load_classification_model(device, model_type="q", num_classes=6)
 
+# b) T5 tokenizer + modelini yükle (Tbai-DPA-1-0 repo’su)
 t5_tokenizer, t5_model = load_t5_model(device)
 
+# c) Performans metriklerini bir kez hesapla (opsiyonel)
 perf_metrics_f = calculate_performance_metrics(cnn_model_f, device)
 perf_metrics_c = calculate_performance_metrics(cnn_model_c, device)
 perf_metrics_q = calculate_performance_metrics(cnn_model_q, device)
 
+
+# ------------------------------------------------------------
+# 2) SABİT VERİLER
+# ------------------------------------------------------------
 class_names_en = [
     "Alzheimer Disease",
     "Mild Alzheimer Risk",
@@ -46,21 +59,30 @@ en2tr = {
     "Parkinson Disease": "Parkinson Hastalığı"
 }
 
+
+# ------------------------------------------------------------
+# 3) GRADIO PREDİKTE FONKSİYONU
+# ------------------------------------------------------------
 def gradio_predict(image, model_type, question):
+    """
+    - image: PIL Image (kullanıcının yüklediği)
+    - model_type: "f", "c" veya "q"
+    - question: isteğe bağlı metin
+    """
+    # A) Seçilen CNN modelini al
     if model_type == "f":
         cnn_model = cnn_model_f
-        perf_metrics = perf_metrics_f
     elif model_type == "c":
         cnn_model = cnn_model_c
-        perf_metrics = perf_metrics_c
     else:
         cnn_model = cnn_model_q
-        perf_metrics = perf_metrics_q
 
+    # B) Görüntü sınıflandırma
     idx, conf, inp_tensor, all_probs = predict_image(cnn_model, image, device)
     pred_en = class_names_en[idx]
     pred_tr = en2tr[pred_en]
 
+    # C) T5 yorumu üret (eğer soru varsa prompt’a ekleyelim)
     if not question or question.strip() == "":
         comment = generate_comment_turkce(t5_tokenizer, t5_model, pred_tr, device)
     else:
@@ -83,12 +105,14 @@ def gradio_predict(image, model_type, question):
         )
         comment = t5_tokenizer.decode(out_ids[0], skip_special_tokens=True)
 
+    # D) Görüntüyü normalize edip numpy array’e dönüştür
     inp_np = inp_tensor.squeeze(0).permute(1, 2, 0).cpu().numpy()
     mean = np.array([0.485, 0.456, 0.406])
     std = np.array([0.229, 0.224, 0.225])
     img_show = inp_np * std + mean
     img_show = np.clip(img_show, 0, 1)
 
+    # E) Ekrana dönecek metinler
     tahmin_metni = f"Tahmin: {pred_en} — {conf:.2f}%"
 
     return img_show, tahmin_metni, comment