jung-geun/PSO
1
0
Fork 0
mirror of https://github.com/jung-geun/PSO.git synced 2025-12-19 20:44:39 +09:00
Code Issues Packages Projects Releases Wiki Activity

23-11-01

Browse Source 
tensorboard 선택 시 자동으로 프로세스 실행
비어있는 포트를 자동으로 탐색하여 오픈
이전 최적해와 비교하여 관성치를 높게 주는 방법을 일시 폐기
digits 테스트 추가
tensorboard 자동 설치 추가
This commit is contained in:
jung-geun
2023-11-01 23:40:31 +09:00
parent 389027409d
commit 80695f304d
9 changed files with 218 additions and 113 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
pso/__init__.py
View File

@@ -1,7 +1,9 @@
from .optimizer import Optimizer as optimizer
from .particle import Particle as particle
__version__ = "1.0.3"
__version__ = "1.0.4"
print("pso2keras version : " + __version__)
__all__ = [
"optimizer",

115
pso/optimizer.py
View File

@@ -1,13 +1,17 @@
import gc
import json
import os
import socket
import subprocess
import sys
from datetime import datetime
import numpy as np
import tensorflow as tf
from tensorboard.plugins.hparams import api as hp
from tensorflow import keras
from tqdm.auto import tqdm
import atexit
from .particle import Particle
@@ -21,6 +25,14 @@ if gpus:
os.environ["TF_CPP_MIN_LOG_LEVEL"] = "2"
def find_free_port():
sock = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
sock.bind(("localhost", 0))
port = sock.getsockname()[1]
sock.close()
return port
class Optimizer:
"""
particle swarm optimization
@@ -33,9 +45,9 @@ class Optimizer:
loss: any = None,
n_particles: int = None,
c0: float = 0.5,
c1: float = 1.5,
w_min: float = 0.5,
w_max: float = 1.5,
c1: float = 0.3,
w_min: float = 0.2,
w_max: float = 0.9,
negative_swarm: float = 0,
mutation_swarm: float = 0,
np_seed: int = None,
@@ -94,11 +106,7 @@ class Optimizer:
if random_state is not None:
np.random.set_state(random_state)
model.compile(
loss=loss,
optimizer="adam",
metrics=["accuracy", "mse"]
)
model.compile(loss=loss, optimizer="adam", metrics=["accuracy", "mse"])
self.model = model # 모델 구조
self.loss = loss # 손실함수
self.n_particles = n_particles # 파티클 개수
@@ -124,11 +132,12 @@ class Optimizer:
print(f"start running time : {self.day}")
for i in tqdm(range(self.n_particles), desc="Initializing Particles"):
self.particles[i] = Particle(
model,
self.loss,
negative=True if i < self.negative_swarm * self.n_particles else False,
negative=True
if i < self.negative_swarm * self.n_particles
else False,
mutation=self.mutation_swarm,
converge_reset=convergence_reset,
converge_reset_patience=convergence_reset_patience,
@@ -308,8 +317,9 @@ class Optimizer:
def __getBatchSlice(self, batch_size):
return list(
tf.data.Dataset.from_tensor_slices(
(self.x, self.y)).shuffle(len(self.x)).batch(batch_size)
tf.data.Dataset.from_tensor_slices((self.x, self.y))
.shuffle(len(self.x))
.batch(batch_size)
)
def getDataset(self):
@@ -353,7 +363,13 @@ class Optimizer:
raise ValueError("x, y shape error")
if log not in [0, 1, 2]:
raise ValueError("log not in [0, 1, 2]")
raise ValueError(
"""log not in [0, 1, 2]
0 : log 기록 안함
1 : csv
2 : tensorboard
"""
)
if renewal not in ["acc", "loss", "mse"]:
raise ValueError("renewal not in ['acc', 'loss', 'mse']")
@@ -394,7 +410,19 @@ class Optimizer:
self.train_summary_writer[i] = tf.summary.create_file_writer(
train_log_dir + f"/{i}"
)
port = find_free_port()
tensorboard_precess = subprocess.Popen(
[
"tensorboard",
"--logdir",
self.log_path,
"--port",
str(port),
]
)
tensorboard_url = f"http://localhost:{port}"
print(f"tensorboard url : {tensorboard_url}")
atexit.register(tensorboard_precess.kill)
elif check_point is not None or log == 1:
if not os.path.exists(self.log_path):
os.makedirs(self.log_path, exist_ok=True)
@@ -408,7 +436,7 @@ class Optimizer:
model_.compile(
loss=self.loss,
optimizer="adam",
metrics=["accuracy", "mse"]
metrics=["accuracy", "mse"],
)
model_.fit(x, y, epochs=1, verbose=0)
score = model_.evaluate(x, y, verbose=1)
@@ -463,8 +491,7 @@ class Optimizer:
if dispersion:
ts = weight_min + np.random.rand() * (weight_max - weight_min)
g_, g_sh, g_len = self._encode(
Particle.g_best_weights)
g_, g_sh, g_len = self._encode(Particle.g_best_weights)
decrement = (epochs - epoch + 1) / epochs
g_ = (1 - decrement) * g_ + decrement * ts
g_best = self._decode_(g_, g_sh, g_len)
@@ -472,8 +499,7 @@ class Optimizer:
if empirical_balance:
if np.random.rand() < np.exp(-(epoch) / epochs):
w_p_ = self._f(
x_batch, y_batch, self.particles[i].get_best_weights(
)
x_batch, y_batch, self.particles[i].get_best_weights()
)
w_g_ = self._f(x_batch, y_batch, g_best)
w_p = w_p_ / (w_p_ + w_g_)
@@ -491,9 +517,7 @@ class Optimizer:
1
/ (
self.n_particles
* np.linalg.norm(
weight_min - weight_max
)
* np.linalg.norm(weight_min - weight_max)
)
* sigma_post
)
@@ -543,9 +567,7 @@ class Optimizer:
if log == 2:
with self.train_summary_writer[i].as_default():
tf.summary.scalar("loss", score[0], step=epoch + 1)
tf.summary.scalar(
"accuracy", score[1], step=epoch + 1
)
tf.summary.scalar("accuracy", score[1], step=epoch + 1)
tf.summary.scalar("mse", score[2], step=epoch + 1)
if renewal == "loss":
@@ -555,6 +577,11 @@ class Optimizer:
min_loss, max_acc, min_mse = score
best_particle_index = i
elif score[0] == min_loss:
if score[1] > max_acc:
min_loss, max_acc, min_mse = score
best_particle_index = i
elif renewal == "acc":
# 최고 점수 보다 높거나 같을 경우
if score[1] > max_acc:
@@ -562,14 +589,22 @@ class Optimizer:
min_loss, max_acc, min_mse = score
best_particle_index = i
elif score[1] == max_acc:
if score[2] < min_mse:
min_loss, max_acc, min_mse = score
best_particle_index = i
elif renewal == "mse":
if score[2] < min_mse:
min_loss, max_acc, min_mse = score
best_particle_index = i
elif score[2] == min_mse:
if score[1] > max_acc:
min_loss, max_acc, min_mse = score
best_particle_index = i
particle_sum += score[1]
if log == 1:
@@ -587,34 +622,28 @@ class Optimizer:
if renewal == "loss":
if min_loss <= Particle.g_best_score[0]:
if min_loss < Particle.g_best_score[0]:
self.particles[best_particle_index].update_global_best(
)
self.particles[best_particle_index].update_global_best()
else:
if max_acc > Particle.g_best_score[1]:
self.particles[best_particle_index].update_global_best(
)
self.particles[best_particle_index].update_global_best()
elif renewal == "acc":
if max_acc >= Particle.g_best_score[1]:
# 최고 점수 보다 높을 경우
if max_acc > Particle.g_best_score[1]:
# 최고 점수 갱신
self.particles[best_particle_index].update_global_best(
)
self.particles[best_particle_index].update_global_best()
# 최고 점수 와 같을 경우
else:
# 최저 loss 보다 낮을 경우
if min_loss < Particle.g_best_score[0]:
self.particles[best_particle_index].update_global_best(
)
self.particles[best_particle_index].update_global_best()
elif renewal == "mse":
if min_mse <= Particle.g_best_score[2]:
if min_mse < Particle.g_best_score[2]:
self.particles[best_particle_index].update_global_best(
)
self.particles[best_particle_index].update_global_best()
else:
if max_acc > Particle.g_best_score[1]:
self.particles[best_particle_index].update_global_best(
)
self.particles[best_particle_index].update_global_best()
# 최고 점수 갱신
epochs_pbar.set_description(
f"best - loss: {Particle.g_best_score[0]:.4f} - acc: {Particle.g_best_score[1]:.4f} - mse: {Particle.g_best_score[2]:.4f}"
@@ -623,9 +652,12 @@ class Optimizer:
if check_point is not None:
if epoch % check_point == 0:
os.makedirs(
f"./logs/{log_name}/{self.day}", exist_ok=True)
f"./logs/{log_name}/{self.day}",
exist_ok=True,
)
self._check_point_save(
f"./logs/{log_name}/{self.day}/ckpt-{epoch}")
f"./logs/{log_name}/{self.day}/ckpt-{epoch}"
)
tf.keras.backend.reset_uids()
tf.keras.backend.clear_session()
@@ -661,7 +693,7 @@ class Optimizer:
model.compile(
loss=self.loss,
optimizer="adam",
metrics=["accuracy", "mse"]
metrics=["accuracy", "mse"],
)
return model
@@ -741,8 +773,7 @@ class Optimizer:
x, y = valid_data
model = self.get_best_model()
score = model.evaluate(x, y, verbose=1)
print(
f"model score - loss: {score[0]} - acc: {score[1]} - mse: {score[2]}")
print(f"model score - loss: {score[0]} - acc: {score[1]} - mse: {score[2]}")
model.save(
f"./{self.log_path}/model_{score[0 if self.renewal == 'loss' else 1 if self.renewal == 'acc' else 2 ]}.h5"
)

56
pso/particle.py
View File

@@ -54,11 +54,11 @@ class Particle:
self.__reset_particle()
self.best_weights = self.model.get_weights()
self.before_best = self.model.get_weights()
# self.before_best = self.model.get_weights()
self.negative = negative
self.mutation = mutation
self.best_score = [np.inf, 0, np.inf]
self.before_w = 0
# self.before_w = 0
self.score_history = []
self.converge_reset = converge_reset
self.converge_reset_patience = converge_reset_patience
@@ -196,7 +196,7 @@ class Particle:
del i_w_, i_s, i_l
self.score_history = []
def _update_velocity(self, local_rate, global_rate, w):
def _velocity_calculation(self, local_rate, global_rate, w):
"""
현재 속도 업데이트
@@ -209,20 +209,20 @@ class Particle:
encode_v, v_sh, v_len = self._encode(weights=self.velocities)
encode_p, p_sh, p_len = self._encode(weights=self.best_weights)
encode_g, g_sh, g_len = self._encode(weights=Particle.g_best_weights)
encode_before, before_sh, before_len = self._encode(
weights=self.before_best
)
# encode_before, before_sh, before_len = self._encode(
# weights=self.before_best
# )
r_0 = np.random.rand()
r_1 = np.random.rand()
# 이전 전역 최적해와 현재 전역 최적해가 다르면 관성을 순간적으로 증가 - 값이 바뀔 경우 기존 관성을 특정 기간동안 유지
if not np.array_equal(encode_before, encode_g, equal_nan=True):
# if not np.array_equal(encode_before, encode_g, equal_nan=True):
# 이전 가중치 중요도의 1.5 배로 관성을 증가
self.before_w = w * 0.5
w = w + self.before_w
else:
self.before_w *= 0.75
w = w + self.before_w
# self.before_w = w * 0.5
# w = w + self.before_w
# else:
# self.before_w *= 0.75
# w = w + self.before_w
if self.negative:
# 지역 최적해와 전역 최적해를 음수로 사용하여 전역 탐색을 유도
@@ -251,10 +251,9 @@ class Particle:
del encode_v, v_sh, v_len
del encode_p, p_sh, p_len
del encode_g, g_sh, g_len
del encode_before, before_sh, before_len
del r_0, r_1
def _update_velocity_w(self, local_rate, global_rate, w, w_p, w_g):
def _velocity_calculation_w(self, local_rate, global_rate, w, w_p, w_g):
"""
현재 속도 업데이트
기본 업데이트의 변형으로 지역 최적해와 전역 최적해를 분산시켜 조기 수렴을 방지
@@ -270,18 +269,18 @@ class Particle:
encode_v, v_sh, v_len = self._encode(weights=self.velocities)
encode_p, p_sh, p_len = self._encode(weights=self.best_weights)
encode_g, g_sh, g_len = self._encode(weights=Particle.g_best_weights)
encode_before, before_sh, before_len = self._encode(
weights=self.before_best
)
# encode_before, before_sh, before_len = self._encode(
# weights=self.before_best
# )
r_0 = np.random.rand()
r_1 = np.random.rand()
if not np.array_equal(encode_before, encode_g, equal_nan=True):
self.before_w = w * 0.5
w = w + self.before_w
else:
self.before_w *= 0.75
w = w + self.before_w
# if not np.array_equal(encode_before, encode_g, equal_nan=True):
# self.before_w = w * 0.5
# w = w + self.before_w
# else:
# self.before_w *= 0.75
# w = w + self.before_w
if self.negative:
new_v = (
@@ -306,10 +305,9 @@ class Particle:
del encode_v, v_sh, v_len
del encode_p, p_sh, p_len
del encode_g, g_sh, g_len
del encode_before, before_sh, before_len
del r_0, r_1
def _update_weights(self):
def _position_update(self):
"""
가중치 업데이트
"""
@@ -337,8 +335,8 @@ class Particle:
Returns:
list: 현재 파티클의 점수
"""
self._update_velocity(local_rate, global_rate, w)
self._update_weights()
self._velocity_calculation(local_rate, global_rate, w)
self._position_update()
score = self.get_score(x, y, renewal)
@@ -385,8 +383,8 @@ class Particle:
Returns:
float: 현재 파티클의 점수
"""
self._update_velocity_w(local_rate, global_rate, w, w_p, w_g)
self._update_weights()
self._velocity_calculation_w(local_rate, global_rate, w, w_p, w_g)
self._position_update()
score = self.get_score(x, y, renewal)