def __init__(self, problem, run, evaluator, **kwargs):
_args = vars(self.parse_args(''))
kwargs['problem'] = problem
kwargs['run'] = run
kwargs['evaluator'] = evaluator
_args.update(kwargs)
_args['problem'] = problem
_args['run'] = run
self.args = Namespace(**_args)
self.problem = util.generic_loader(problem, 'Problem')
self.run_func = util.generic_loader(run, 'run')
logger.info('Evaluator will execute the function: ' + run)
if kwargs.get('cache_key') is None:
self.evaluator = Evaluator.create(self.run_func, method=evaluator)
else:
self.evaluator = Evaluator.create(self.run_func,
method=evaluator,
cache_key=kwargs['cache_key'])
self.num_workers = self.evaluator.num_workers
logger.info(f'Options: ' + pformat(self.args.__dict__, indent=4))
logger.info('Hyperparameter space definition: ' +
pformat(self.problem.space, indent=4))
logger.info(f'Created {self.args.evaluator} evaluator')
logger.info(f'Evaluator: num_workers is {self.num_workers}')
def __init__(self, comm, problem, run, evaluator, **kwargs):
"""Constructor."""
self.comm = comm
self.rank = comm.Get_rank()
self.problem = util.generic_loader(problem, "Problem")
self.run_func = util.generic_loader(run, "run")
self.evaluator = Evaluator.create(self.run_func, method=evaluator)
def setUp(self):
self.ev = Evaluator.create(run, cache_key=key, method="balsam")
self.AWAIT_TIMEOUT = 30
self.launcher = subprocess.Popen(
"balsam launcher --job-mode=serial --consume-all --num-transition=1",
shell=True,
)
def __init__(self, problem, fbest, evaluator, **kwargs):
if MPI is None:
self.free_workers = 1
else:
nranks = MPI.COMM_WORLD.Get_size()
if evaluator == 'balsam':
balsam_launcher_nodes = int(
os.environ.get('BALSAM_LAUNCHER_NODES', 1))
deephyper_workers_per_node = int(
os.environ.get('DEEPHYPER_WORKERS_PER_NODE', 1))
n_free_nodes = balsam_launcher_nodes - nranks # Number of free nodes
self.free_workers = n_free_nodes * \
deephyper_workers_per_node # Number of free workers
else:
self.free_workers = 1
_args = vars(self.parse_args(''))
kwargs['problem'] = problem
kwargs['p_f_best'] = fbest
kwargs['evaluator'] = evaluator
_args.update(kwargs)
self.args = Namespace(**_args)
self.problem = util.generic_loader(problem, 'Problem')
if kwargs.get('cache_key') is None:
self.evaluator = Evaluator.create(run_function=train,
method=evaluator)
else:
self.evaluator = Evaluator.create(run_function=train,
method=evaluator,
cache_key=kwargs['cache_key'])
self.num_workers = self.evaluator.num_workers
logger.info(f'Options: ' + pformat(self.args.__dict__, indent=4))
logger.info('Problem definition: ' +
pformat(self.problem.space, indent=4))
logger.info(f'Created {self.args.evaluator} evaluator')
logger.info(f'Evaluator: num_workers is {self.num_workers}')
def __init__(self, problem, fbest, evaluator, **kwargs):
if MPI is None:
self.free_workers = 1
else:
nranks = MPI.COMM_WORLD.Get_size()
if evaluator == "balsam":
balsam_launcher_nodes = int(
os.environ.get("BALSAM_LAUNCHER_NODES", 1))
deephyper_workers_per_node = int(
os.environ.get("DEEPHYPER_WORKERS_PER_NODE", 1))
n_free_nodes = balsam_launcher_nodes - nranks # Number of free nodes
self.free_workers = (n_free_nodes * deephyper_workers_per_node
) # Number of free workers
else:
self.free_workers = 1
_args = vars(self.parse_args(""))
kwargs["problem"] = problem
kwargs["p_f_best"] = fbest
kwargs["evaluator"] = evaluator
_args.update(kwargs)
self.args = Namespace(**_args)
self.problem = util.generic_loader(problem, "Problem")
if kwargs.get("cache_key") is None:
self.evaluator = Evaluator.create(run_function=train,
method=evaluator)
else:
self.evaluator = Evaluator.create(run_function=train,
method=evaluator,
cache_key=kwargs["cache_key"])
self.num_workers = self.evaluator.num_workers
logger.info(f"Options: " + pformat(self.args.__dict__, indent=4))
logger.info("Problem definition: " +
pformat(self.problem.space, indent=4))
logger.info(f"Created {self.args.evaluator} evaluator")
logger.info(f"Evaluator: num_workers is {self.num_workers}")
def __init__(
self,
problem: str,
run: str,
evaluator: str,
max_evals: int = 1000000,
seed: int = None,
num_nodes_master: int = 1,
num_workers: int = None,
log_dir: int = None,
**kwargs,
):
kwargs["problem"] = problem
kwargs["run"] = run
kwargs["evaluator"] = evaluator
kwargs["max_evals"] = max_evals # * For retro compatibility
kwargs["seed"] = seed
# Loading problem instance and run function
self.problem = util.generic_loader(problem, "Problem")
if self.problem.seed == None:
self.problem.seed = seed
else:
kwargs["seed"] = self.problem.seed
self.run_func = util.generic_loader(run, "run")
notice = f"Maximizing the return value of function: {run}"
logger.info(notice)
util.banner(notice)
self.evaluator = Evaluator.create(
self.run_func,
method=evaluator,
num_nodes_master=num_nodes_master,
num_workers=num_workers,
**kwargs,
)
self.num_workers = self.evaluator.num_workers
self.max_evals = max_evals
self.log_dir = os.getcwd() if log_dir is None else log_dir
# set the random seed
np.random.seed(self.problem.seed)
logger.info(f"Options: " + pformat(kwargs, indent=4))
logger.info("Hyperparameter space definition: " +
pformat(self.problem.space, indent=4))
logger.info(f"Created {evaluator} evaluator")
logger.info(f"Evaluator: num_workers is {self.num_workers}")
def _on_done(job): # def _on_done(job, process_data):
output = job.read_file_in_workdir(f'{job.name}.out')
# process_data(job)
#args = job.args
#args = args.replace("\'", "")
# with open('test.json', 'w') as f:
# f.write(args)
#
# with open('test.json', 'r') as f:
# args = json.load(f)
output = Evaluator._parse(output)
#job.data['reward'] = output
#job.data['arch_seq'] = args['arch_seq']
#job.data['id_worker'] = args['w']
# job.save()
return output
def __init__(self,
problem: str,
run: str,
evaluator: str,
max_evals: int = 100,
seed: int = None,
**kwargs):
kwargs['problem'] = problem
kwargs['run'] = run
kwargs['evaluator'] = evaluator
kwargs['max_evals'] = max_evals # * For retro compatibility
kwargs['seed'] = seed
self.problem = util.generic_loader(problem, 'Problem')
if self.problem.seed == None:
self.problem.seed = seed
else:
kwargs['seed'] = self.problem.seed
self.run_func = util.generic_loader(run, 'run')
notice = f'Maximizing the return value of function: {run}'
logger.info(notice)
util.banner(notice)
self.evaluator = Evaluator.create(self.run_func,
method=evaluator,
**kwargs)
self.num_workers = self.evaluator.num_workers
self.max_evals = max_evals
# set the random seed
np.random.seed(self.problem.seed)
logger.info(f'Options: ' + pformat(kwargs, indent=4))
logger.info('Hyperparameter space definition: ' +
pformat(self.problem.space, indent=4))
logger.info(f'Created {evaluator} evaluator')
logger.info(f'Evaluator: num_workers is {self.num_workers}')
def setUp(self):
self.ev = Evaluator.create(run, cache_key=key, method='local-lite')
self.AWAIT_TIMEOUT = 8
def _on_done(job):
if "dh_objective" in job.data:
return job.data["dh_objective"]
output = job.read_file_in_workdir(f"{job.name}.out")
output = Evaluator._parse(output)
return output