Add better support for multiprocessing of usage: optuna [--version] […

…-v | -q] [--log-file LOG_FILE] [-h] [--debug]

              [--config CONFIG] [--storage STORAGE]

optional arguments:
  --version            show program's version number and exit
  -v, --verbose        Increase verbosity of output. Can be repeated.
  -q, --quiet          Suppress output except warnings and errors.
  --log-file LOG_FILE  Specify a file to log output. Disabled by default.
  -h, --help           Show help message and exit.
  --debug              Show tracebacks on errors.
  --config CONFIG      Config file path. (default=$HOME/.optuna.yml)
  --storage STORAGE    DB URL. (e.g. sqlite:///example.db)

Commands:
  complete       print bash completion command (cliff)
  create-study
  dashboard
  help           print detailed help for another command (cliff)
  storage upgrade
  studies        Command base class for providing a list of data as output.
  study optimize
  study set-user-attr   optimizing.

lib/RLTrader.py

@@ -120,14 +120,11 @@ def optimize_params(self, trial, n_prune_evals_per_trial: int = 2, n_tests_per_e
 
         del test_provider
 
-        n_envs = min(self.n_envs, 4)
-        n_minibatches = self.n_minibatches if n_envs % self.n_minibatches == 0 else n_envs
-
-        train_env = SubprocVecEnv([make_env(train_provider, i) for i in range(n_envs)])
-        validation_env = SubprocVecEnv([make_env(validation_provider, i) for i in range(n_envs)])
+        train_env = SubprocVecEnv([make_env(train_provider, i) for i in range(1)])
+        validation_env = SubprocVecEnv([make_env(validation_provider, i) for i in range(1)])
 
         model_params = self.optimize_agent_params(trial)
-        model = self.Model(self.Policy, train_env, verbose=self.model_verbose, nminibatches=n_minibatches,
+        model = self.Model(self.Policy, train_env, verbose=self.model_verbose, nminibatches=1,
                            tensorboard_log=self.tensorboard_path, **model_params)
 
         last_reward = -np.finfo(np.float16).max

optimize.py

@@ -1,11 +1,30 @@
 import numpy as np
-
+import multiprocessing
 from lib.RLTrader import RLTrader
 
 np.warnings.filterwarnings('ignore')
 
-if __name__ == '__main__':
-    trader = RLTrader()
+
+def optimize_code(params):
+    trader = RLTrader(**params)
 
     trader.optimize()
+
+
+if __name__ == '__main__':
+    n_process = multiprocessing.cpu_count()
+    params = {}
+
+    process = []
+    for i in range(n_process):
+        process.append(multiprocessing.Process(target=optimize_code, args=(params,)))
+
+    for p in process:
+        p.start()
+
+    for p in process:
+        p.join()
+
+    trader = RLTrader(**params)
+
     trader.train(test_trained_model=True, render_trained_model=True)