Rename zero_shot_classification.py to app.py

zero_shot_classification.py → app.py

@@ -57,7 +57,7 @@ LABELS_MAP = ["Bat (baseball)", "Bat (mammal)",
 
 """
 
-model_key  = "CLIP-large"
+# model_key  = "CLIP-large"
 
 # Load model (cache for speed)
 if model_key not in MODEL_CACHE:
@@ -73,8 +73,8 @@ output = classifier(
     candidate_labels    = CANDIDATE_LABELS,
     hypothesis_template = "This image shows {}")
 
-print("\n\n=============================================================================")
-print(f"\nPrediction: This image shows {output[0]["label"]} | Confidence (probability): {100*output[0]["score"]: .1f}%")
+# print("\n\n=============================================================================")
+# print(f"\nPrediction: This image shows {output[0]["label"]} | Confidence (probability): {100*output[0]["score"]: .1f}%")
 
 def run_classifer(model_key, image_path, prob_threshold = None):
     # model_key: name of backbone zero-shot-image-classification model to use
@@ -107,17 +107,15 @@ def run_classifer(model_key, image_path, prob_threshold = None):
 
     return predicted_label_str, prob_dict
 
-# example run
-model_key  = "CLIP-large"
-BASE_DIR   = '/content/drive/MyDrive/ML Projects/Zero-shot Image Classification/Images'
-image_path = os.path.join(BASE_DIR, 'Nail2_1.png')
+# # example run
+# model_key  = "CLIP-large"
+# BASE_DIR   = '/content/drive/MyDrive/ML Projects/Zero-shot Image Classification/Images'
+# image_path = os.path.join(BASE_DIR, 'Nail2_1.png')
 
-predicted_label_str, prob_dict = run_classifer(model_key, image_path, prob_threshold = 0.4)
-print("\n\n=============================================================================")
-# print(f"\nPrediction: {predicted_label_str} | Confidence (probability): {100*output[0]['score']:.1f}%")
-print(f"\nPrediction: {predicted_label_str}")
-
-prob_dict
+# predicted_label_str, prob_dict = run_classifer(model_key, image_path, prob_threshold = 0.4)
+# print("\n\n=============================================================================")
+# # print(f"\nPrediction: {predicted_label_str} | Confidence (probability): {100*output[0]['score']:.1f}%")
+# print(f"\nPrediction: {predicted_label_str}")
 
 """### Gradio App