23-06-24

패키지 호출 단순 수정

mnist.py

@@ -1,30 +1,28 @@
 # %%
 import os
+
 os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'
 
 import tensorflow as tf
+
 tf.random.set_seed(777)  # for reproducibility
 
 import numpy as np
+
 np.random.seed(777)
 
-from tensorflow import keras
-from keras.datasets import mnist
-from keras.models import Sequential
-from keras.layers import Dense, Dropout, Flatten
-from keras.layers import Conv2D, MaxPooling2D
-from keras import backend as K
+import gc
+from datetime import date
 
+from keras import backend as K
+from keras.datasets import mnist
+from keras.layers import Conv2D, Dense, Dropout, Flatten, MaxPooling2D
+from keras.models import Sequential
 # from pso_tf import PSO
 from pso import Optimizer
-# from optimizer import Optimizer
-
-
-from datetime import date
+from tensorflow import keras
 from tqdm import tqdm
 
-import gc
-
 # print(tf.__version__)
 # print(tf.config.list_physical_devices())
 # print(f"Num GPUs Available: {len(tf.config.list_physical_devices('GPU'))}")
@@ -74,12 +72,13 @@ if __name__ == "__main__":
         pso_mnist = Optimizer(
             model,
             loss=loss[0], 
-            n_particles=50,
+            n_particles=75,
             c0=0.35, 
             c1=0.8, 
             w_min=0.7,
             w_max=1.15,
-            negative_swarm=0.25
+            negative_swarm=0.25,
+            momentun_swarm=0.25,
             )
 
         best_score = pso_mnist.fit(
@@ -88,7 +87,7 @@ if __name__ == "__main__":
             epochs=200,
             save=True,
             save_path="./result/mnist", 
-            renewal="acc", 
+            renewal="loss", 
             empirical_balance=False,
             Dispersion=False, 
             check_point=25

pso/optimizer.py

@@ -1,19 +1,18 @@
+import gc
+import json
 import os
 import sys
-
-import tensorflow as tf
-from tensorflow import keras
+from datetime import datetime
 
 import numpy as np
+import tensorflow as tf
+from pso.particle import Particle
+from tensorflow import keras
+from tqdm import tqdm
+
 # import cupy as cp
 
-from tqdm import tqdm
-from datetime import datetime
-import json
-import gc
-from copy import copy, deepcopy
 
-from pso.particle import Particle
 
 gpus = tf.config.experimental.list_physical_devices("GPU")
 if gpus:
@@ -35,6 +34,7 @@ class Optimizer:
         w_min=0.5,
         w_max=1.5,
         negative_swarm: float = 0,
+        momentun_swarm: float = 0,
     ):
         """
         Args:
@@ -46,6 +46,7 @@ class Optimizer:
             w_min (float): 최소 관성 수치
             w_max (float): 최대 관성 수치
             nefative_swarm (float): 최적해와 반대로 이동할 파티클 비율 - 0 ~ 1 사이의 값
+            momentun_swarm (float): 관성을 추가로 사용할 파티클 비율 - 0 ~ 1 사이의 값
         """
         self.model = model  # 모델 구조
         self.loss = loss  # 손실함수
@@ -56,6 +57,7 @@ class Optimizer:
         self.w_min = w_min  # 최소 관성 수치
         self.w_max = w_max  # 최대 관성 수치
         self.negative_swarm = negative_swarm  # 최적해와 반대로 이동할 파티클 비율 - 0 ~ 1 사이의 값
+        self.momentun_swarm = momentun_swarm  # 관성을 추가로 사용할 파티클 비율 - 0 ~ 1 사이의 값
         self.g_best_score = [0 , np.inf]  # 최고 점수 - 시작은 0으로 초기화
         self.g_best = None  # 최고 점수를 받은 가중치
         self.g_best_ = None  # 최고 점수를 받은 가중치 - 값의 분산을 위한 변수
@@ -68,10 +70,8 @@ class Optimizer:
             w_ = np.random.uniform(-1.5, 1.5, len(w_))
             m.set_weights(self._decode(w_, sh_, len_))
             m.compile(loss=self.loss, optimizer="sgd", metrics=["accuracy"])
-            if i < negative_swarm * self.n_particles:
-                self.particles[i] = Particle(m, loss, negative=True)
-            else:
-                self.particles[i] = Particle(m, loss, negative=False)
+            self.particles[i] = Particle(m, loss, negative=True if i < negative_swarm * self.n_particles else False, momentun=True if i > self.n_particles * (1 - self.momentun_swarm) else False)
+                
         gc.collect()
         
     def __del__(self):

pso/particle.py

@@ -1,15 +1,16 @@
-import tensorflow as tf
-from tensorflow import keras
+import gc
 
 # import cupy as cp
 import numpy as np
-import gc
+import tensorflow as tf
+from tensorflow import keras
+
 
 class Particle:
     """
     Particle Swarm Optimization의 Particle을 구현한 클래스
     """
-    def __init__(self, model: keras.models, loss, negative: bool = False|True):
+    def __init__(self, model: keras.models, loss, negative: bool = False, momentun: bool = False):
         """
         Args:
             model (keras.models): 학습 및 검증을 위한 모델
@@ -23,6 +24,7 @@ class Particle:
         i_w_ = np.random.rand(len(i_w_)) / 2 - 0.25
         self.velocities = self._decode(i_w_, s_, l_)
         self.negative = negative
+        self.momentun = momentun
         self.best_score = 0
         self.best_weights = init_weights
 
@@ -146,6 +148,8 @@ class Particle:
                 + local_rate * r0 * (encode_p - encode_w)
                 + global_rate * r1 * (encode_g - encode_w)
             )
+        if self.momentun:
+            new_v += 0.5 * encode_v
         self.velocities = self._decode(new_v, w_sh, w_len)
         del encode_w, w_sh, w_len
         del encode_v, v_sh, v_len
@@ -184,6 +188,8 @@ class Particle:
                 + local_rate * r0 * (w_p * encode_p - encode_w)
                 + global_rate * r1 * (w_g * encode_g - encode_w)
             )
+        if self.momentun:
+            new_v += 0.5 * encode_v
         self.velocities = self._decode(new_v, w_sh, w_len)
         del encode_w, w_sh, w_len
         del encode_v, v_sh, v_len