def start(self, port: int = 8080, debug: bool = False) -> None:
"""
Start the experiment in background.
This method will raise exception on failure.
If it returns, the experiment should have been successfully started.
Parameters
----------
port
The port of web UI.
debug
Whether to start in debug mode.
"""
atexit.register(self.stop)
self.id = management.generate_experiment_id()
if self.config.experiment_working_directory is not None:
log_dir = Path(self.config.experiment_working_directory, self.id,
'log')
else:
log_dir = Path.home() / f'nni-experiments/{self.id}/log'
nni.runtime.log.start_experiment_log(self.id, log_dir, debug)
self._proc, self._pipe = launcher.start_experiment_retiarii(
self.id, self.config, port, debug)
assert self._proc is not None
assert self._pipe is not None
self.port = port # port will be None if start up failed
# dispatcher must be launched after pipe initialized
# the logic to launch dispatcher in background should be refactored into dispatcher api
self._dispatcher = self._create_dispatcher()
self._dispatcher_thread = Thread(target=self._dispatcher.run)
self._dispatcher_thread.start()
ips = [self.config.nni_manager_ip]
for interfaces in psutil.net_if_addrs().values():
for interface in interfaces:
if interface.family == socket.AF_INET:
ips.append(interface.address)
ips = [f'http://{ip}:{port}' for ip in ips if ip]
msg = 'Web UI URLs: ' + colorama.Fore.CYAN + ' '.join(
ips) + colorama.Style.RESET_ALL
_logger.info(msg)
exp_status_checker = Thread(target=self._check_exp_status)
exp_status_checker.start()
self._start_strategy()
# TODO: the experiment should be completed, when strategy exits and there is no running job
_logger.info(
'Waiting for experiment to become DONE (you can ctrl+c if there is no running trial jobs)...'
)
exp_status_checker.join()
def start(self, port: int = 8080, debug: bool = False) -> None:
"""
Start the experiment in background.
This method will raise exception on failure.
If it returns, the experiment should have been successfully started.
Parameters
----------
port
The port of web UI.
debug
Whether to start in debug mode.
"""
atexit.register(self.stop)
self.config = self.config.canonical_copy()
# we will probably need a execution engine factory to make this clean and elegant
if self.config.execution_engine == 'base':
from ..execution.base import BaseExecutionEngine
engine = BaseExecutionEngine()
elif self.config.execution_engine == 'cgo':
from ..execution.cgo_engine import CGOExecutionEngine
assert self.config.training_service.platform == 'remote', \
"CGO execution engine currently only supports remote training service"
assert self.config.batch_waiting_time is not None
devices = self._construct_devices()
engine = CGOExecutionEngine(
devices,
max_concurrency=self.config.max_concurrency_cgo,
batch_waiting_time=self.config.batch_waiting_time)
elif self.config.execution_engine == 'py':
from ..execution.python import PurePythonExecutionEngine
engine = PurePythonExecutionEngine()
elif self.config.execution_engine == 'benchmark':
from ..execution.benchmark import BenchmarkExecutionEngine
engine = BenchmarkExecutionEngine(self.config.benchmark)
set_execution_engine(engine)
self.id = management.generate_experiment_id()
if self.config.experiment_working_directory is not None:
log_dir = Path(self.config.experiment_working_directory, self.id,
'log')
else:
log_dir = Path.home() / f'nni-experiments/{self.id}/log'
nni.runtime.log.start_experiment_log(self.id, log_dir, debug)
self._proc, self._pipe = launcher.start_experiment_retiarii(
self.id, self.config, port, debug)
assert self._proc is not None
assert self._pipe is not None
self.port = port # port will be None if start up failed
# dispatcher must be launched after pipe initialized
# the logic to launch dispatcher in background should be refactored into dispatcher api
self._dispatcher = self._create_dispatcher()
self._dispatcher_thread = Thread(target=self._dispatcher.run)
self._dispatcher_thread.start()
ips = [self.config.nni_manager_ip]
for interfaces in psutil.net_if_addrs().values():
for interface in interfaces:
if interface.family == socket.AF_INET:
ips.append(interface.address)
ips = [f'http://{ip}:{port}' for ip in ips if ip]
msg = 'Web UI URLs: ' + colorama.Fore.CYAN + ' '.join(
ips) + colorama.Style.RESET_ALL
_logger.info(msg)
exp_status_checker = Thread(target=self._check_exp_status)
exp_status_checker.start()
self._start_strategy()
# TODO: the experiment should be completed, when strategy exits and there is no running job
_logger.info(
'Waiting for experiment to become DONE (you can ctrl+c if there is no running trial jobs)...'
)
exp_status_checker.join()
