refactor: organize with module and add logs

ai/embedding_util.py

@@ -6,26 +6,28 @@
 from numpy import ndarray
 
 
-def cos_similarity(a: ndarray, b: ndarray) -> float:
-    assert a.ndim == 1, f"The dim of a should be 1, but was {a.ndim}"
-    assert a.shape == b.shape, f"The shape of a and b should be the same, but was {a.shape} and {b.shape}"
-    dot_product = np.dot(a, b)
-    norm_a = np.linalg.norm(a)
-    norm_b = np.linalg.norm(b)
-    return round(dot_product / (norm_a * norm_b), 5)
+class EmbeddingUtil:
+    @classmethod
+    def cos_similarity(cls, a: ndarray, b: ndarray) -> float:
+        assert a.ndim == 1, f"The dim of a should be 1, but was {a.ndim}"
+        assert a.shape == b.shape, f"The shape of a and b should be the same, but was {a.shape} and {b.shape}"
+        dot_product = np.dot(a, b)
+        norm_a = np.linalg.norm(a)
+        norm_b = np.linalg.norm(b)
+        return round(dot_product / (norm_a * norm_b), 5)
 
+    @classmethod
+    def cos_similarity_normalized(cls, a: ndarray, b: ndarray):
+        temp = cls.cos_similarity(a, b)
+        upper_bound = 1
+        lower_bound = -1
+        return (temp - lower_bound) / (upper_bound - lower_bound)
 
-def cos_similarity_normalized(a: ndarray, b: ndarray):
-    temp = cos_similarity(a, b)
-    upper_bound = 1
-    lower_bound = -1
-    return (temp - lower_bound) / (upper_bound - lower_bound)
-
-
-def self_cos_similarty_compare(embedding_list):
-    n = len(embedding_list)
-    res = np.zeros((n, n))
-    for i in range(n):
-        for j in range(n):
-            res[i][j] = cos_similarity_normalized(embedding_list[i], embedding_list[j])
-    return np.round(res, 4)
+    @classmethod
+    def self_cos_similarity_compare(cls, embedding_list):
+        n = len(embedding_list)
+        res = np.zeros((n, n))
+        for i in range(n):
+            for j in range(n):
+                res[i][j] = cls.cos_similarity_normalized(embedding_list[i], embedding_list[j])
+        return np.round(res, 4)

ai/testing_util.py

@@ -10,17 +10,21 @@
 from android_testing_utils.log import my_logger
 
 
-def eliminate_randomness(seed):
-    my_logger.hint(my_logger.LogLevel.WARNING, "TestingUtil", True,
-                   f" #### Eliminate randomness with seed {seed} ####")
-    torch.manual_seed(seed)
-    torch.cuda.manual_seed_all(seed)
-    np.random.seed(seed)
-    random.seed(seed)
-    torch.backends.cudnn.deterministic = True
+class TestingUtil:
+    @classmethod
+    def eliminate_randomness(cls, seed):
+        my_logger.auto_hint(
+            my_logger.LogLevel.WARNING, cls, True, f" #### Eliminate randomness with seed {seed} ####"
+        )
+        torch.manual_seed(seed)
+        torch.cuda.manual_seed_all(seed)
+        np.random.seed(seed)
+        random.seed(seed)
+        torch.backends.cudnn.deterministic = True
 
-
-def model_info(model, input_shape):
-    device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
-    backbone = model.to(device)
-    summary(backbone, input_shape)
+    @classmethod
+    def model_info(cls, model, input_shape):
+        my_logger.auto_hint(my_logger.LogLevel.INFO, cls, True, f" #### Model Summary ####")
+        device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
+        backbone = model.to(device)
+        summary(backbone, input_shape)

monitor/decorator.py

@@ -14,8 +14,10 @@ def wrapper(*args, **kwargs):
         before = time.time()
         res = function(*args, **kwargs)
         after = time.time()
-        my_logger.hint(my_logger.LogLevel.INFO, "TimeCountDecorator", False,
-                       f"Function {function.__name__} cost time: {round(after-before, 4)}s")
+        my_logger.auto_hint(
+            my_logger.LogLevel.INFO, "TimeCountDecorator", False,
+            f"Function {function.__name__} cost time: {round(after-before, 4)}s"
+        )
         return res
     return wrapper
 
@@ -26,8 +28,10 @@ def wrapper(self, *args, **kwargs):
         before = time.time()
         res = function(self, *args, **kwargs)
         after = time.time()
-        my_logger.hint(my_logger.LogLevel.INFO, "TimeCountDecorator", False,
-                       f"Function {function.__name__} cost time: {round(after-before, 2)}s")
+        my_logger.auto_hint(
+            my_logger.LogLevel.INFO, "TimeCountDecorator", False,
+            f"Function {function.__name__} cost time: {round(after-before, 2)}s"
+        )
         return res
     return wrapper
 
@@ -38,7 +42,9 @@ def wrapper(*args, **kwargs):
         before = time.time()
         res = function(*args, **kwargs)
         after = time.time()
-        my_logger.hint(my_logger.LogLevel.INFO, "TimeCountDecorator", False,
-                       f"Function {function.__name__} cost time: {round(after-before, 2)}s")
+        my_logger.auto_hint(
+            my_logger.LogLevel.INFO, "TimeCountDecorator", False,
+            f"Function {function.__name__} cost time: {round(after-before, 2)}s"
+        )
         return res
     return wrapper