Merge pull request #2294 from hroff-1902/fix-skopt-memory3

Fix skopt memory exhaustion
This commit is contained in:
Matthias 2019-09-25 19:55:27 +02:00 committed by GitHub
commit b994f5c273
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 28 additions and 4 deletions

View File

@ -36,6 +36,11 @@ logger = logging.getLogger(__name__)
INITIAL_POINTS = 30
# Keep no more than 2*SKOPT_MODELS_MAX_NUM models
# in the skopt models list
SKOPT_MODELS_MAX_NUM = 10
MAX_LOSS = 100000 # just a big enough number to be bad result in loss optimization
@ -255,12 +260,13 @@ class Hyperopt:
spaces += self.custom_hyperopt.stoploss_space()
return spaces
def generate_optimizer(self, _params: Dict) -> Dict:
def generate_optimizer(self, _params: Dict, iteration=None) -> Dict:
"""
Used Optimize function. Called once per epoch to optimize whatever is configured.
Keep this function as optimized as possible!
"""
params = self.get_args(_params)
if self.has_space('roi'):
self.backtesting.strategy.minimal_roi = \
self.custom_hyperopt.generate_roi_table(params)
@ -342,9 +348,26 @@ class Hyperopt:
random_state=self.config.get('hyperopt_random_state', None)
)
def run_optimizer_parallel(self, parallel, asked) -> List:
def fix_optimizer_models_list(self):
"""
WORKAROUND: Since skopt is not actively supported, this resolves problems with skopt
memory usage, see also: https://github.com/scikit-optimize/scikit-optimize/pull/746
This may cease working when skopt updates if implementation of this intrinsic
part changes.
"""
n = len(self.opt.models) - SKOPT_MODELS_MAX_NUM
# Keep no more than 2*SKOPT_MODELS_MAX_NUM models in the skopt models list,
# remove the old ones. These are actually of no use, the current model
# from the estimator is the only one used in the skopt optimizer.
# Freqtrade code also does not inspect details of the models.
if n >= SKOPT_MODELS_MAX_NUM:
logger.debug(f"Fixing skopt models list, removing {n} old items...")
del self.opt.models[0:n]
def run_optimizer_parallel(self, parallel, asked, i) -> List:
return parallel(delayed(
wrap_non_picklable_objects(self.generate_optimizer))(v) for v in asked)
wrap_non_picklable_objects(self.generate_optimizer))(v, i) for v in asked)
def load_previous_results(self):
""" read trials file if we have one """
@ -405,8 +428,9 @@ class Hyperopt:
EVALS = max(self.total_epochs // jobs, 1)
for i in range(EVALS):
asked = self.opt.ask(n_points=jobs)
f_val = self.run_optimizer_parallel(parallel, asked)
f_val = self.run_optimizer_parallel(parallel, asked, i)
self.opt.tell(asked, [v['loss'] for v in f_val])
self.fix_optimizer_models_list()
for j in range(jobs):
current = i * jobs + j
val = f_val[j]