def choose_next(self, num_config):
if len(self.target_y[self.iterate_r[-1]]) == 0:
configs = sample_configurations(self.config_space, num_config)
self.configs.extend(configs)
return configs
config_candidates = list()
acq_configs = self.get_bo_candidates(num_configs=2 * num_config)
acq_idx = 0
for idx in range(1, 1 + 2 * num_config):
# Like BOHB, sample a fixed percentage of random configurations.
if self.random_configuration_chooser.check(idx):
_config = self.config_space.sample_configuration()
else:
_config = acq_configs[acq_idx]
acq_idx += 1
if _config not in config_candidates:
config_candidates.append(_config)
if len(config_candidates) >= num_config:
break
if len(config_candidates) < num_config:
config_candidates = expand_configurations(config_candidates, self.config_space, num_config)
_config_candidates = []
for config in config_candidates:
if config not in self.configs: # Check if evaluated
_config_candidates.append(config)
self.configs.extend(_config_candidates)
return _config_candidates
def choose_next(self, num_config):
# Sample n configurations according to BOHB strategy.
self.logger.info('Sample %d configs in choose_next. rand_prob is %f.' % (num_config, self.rand_prob))
# get bo configs
# update batchsize each round. random ratio is fixed.
self.config_advisor.batch_size = num_config - int(num_config * self.rand_prob)
bo_configs = self.config_advisor.get_suggestions()
bo_configs = bo_configs[:num_config] # may exceed num_config in initial random sampling
self.logger.info('len bo configs = %d.' % len(bo_configs))
# sample random configs
configs = expand_configurations(bo_configs, self.config_space, num_config)
self.logger.info('len total configs = %d.' % len(configs))
assert len(configs) == num_config
return configs