ARM-software · tihom · Mar 11, 2026 · Mar 11, 2026 · ARM-software · tihom
diff --git a/README.md b/README.md
@@ -396,6 +396,7 @@ Note that if `initial_custom` option is given then `initial_random` is ignored.
 }
 ```
 However, use this option with caution as it could have unintended consequences. 
+- *log_progress*: `True` or `False` (default: `True`). When `True`, Mango logs optimization progress with a tqdm-based progress bar and per-iteration best scores. Set this to `False` to suppress progress logging, which is useful in CI environments or when you want cleaner logs.
 
 
 

diff --git a/development.md b/development.md
@@ -7,4 +7,24 @@ poetry config pypi-token.pypi <api_token>
 > poetry version [major, minor, patch]
 > poetry build
 > poetry publish
+```
+
+### Local development: virtualenv and tests
+
+Create a virtual environment in the project root:
+```shell
+python3 -m venv venv
+```
+
+Activate it from **fish**:
+```shell
+source ./venv/bin/activate.fish
+```
+
+Install development dependencies and run the test suite:
+```shell
+pip install --upgrade pip
+pip install poetry
+poetry install
+pytest
 ```
diff --git a/examples/__init__.py b/examples/__init__.py
diff --git a/examples/classifiers/__init__.py b/examples/classifiers/__init__.py
diff --git a/mango/domain/__init__.py b/mango/domain/__init__.py
diff --git a/mango/metatuner.py b/mango/metatuner.py
@@ -22,6 +22,7 @@ class MetaTuner:
  class Config:
  n_iter: int = 20
  n_init: int = 2
+ log_progress: bool = True
 
  def __init__(self, param_dict_list, objective_list, **kwargs):
  self.param_dict_list = param_dict_list
@@ -30,6 +31,7 @@ def __init__(self, param_dict_list, objective_list, **kwargs):
  # adjustable parameters
  self.num_of_iterations = self.config.n_iter
  self.initial_random = self.config.n_init
+ self.log_progress = self.config.log_progress
 
  # list of GPR for each objective
  self.gpr_list = []
@@ -222,9 +224,13 @@ def runExponentialTuner(self):
  # print(Optimizer_exploration)
 
  # Now run the optimization iterations
- pbar = tqdm(range(self.num_of_iterations))
+ iterator = (
+ tqdm(range(self.num_of_iterations))
+ if self.log_progress
+ else range(self.num_of_iterations)
+ )
 
- for itr in pbar:
+ for itr in iterator:
  # next values of x returned from individual function
  # Values in x are dependent on types of param dict, so using a list
  x_values_list = []
@@ -375,7 +381,8 @@ def runExponentialTuner(self):
  Optimizer_exploration[i] + 1.1 * self.exploration_rate
  )
 
- pbar.set_description(": Best score: %s" % max_val_y)
+ if self.log_progress:
+ tqdm.write(": Best score: %s" % max_val_y)
 
  # print(itr, s_values_list, Optimizer_iteration, Optimizer_exploration, max_val_y)#, Y_dict_array[selected_obj])
 

diff --git a/mango/optimizer/__init__.py b/mango/optimizer/__init__.py
diff --git a/mango/tuner.py b/mango/tuner.py
@@ -45,6 +45,7 @@ class TunerConfig:
  constraint: Callable = None
  param_sampler: Callable = parameter_sampler
  scale_params: bool = False
+ log_progress: bool = True
 
  def __post_init__(self):
  if self.optimizer not in self.valid_optimizers:
@@ -202,8 +203,12 @@ def runBayesianOptimizer(self):
  X_domain_np = self.ds.convert_GP_space(domain_list)
 
  # running the iterations
- pbar = tqdm(range(self.config.num_iteration))
- for _ in pbar:
+ iterator = (
+ tqdm(range(self.config.num_iteration))
+ if self.config.log_progress
+ else range(self.config.num_iteration)
+ )
+ for _ in iterator:
 
  # adding a Minimum exploration to explore independent of UCB
  if random.random() < self.config.exploration:
@@ -284,7 +289,8 @@ def runBayesianOptimizer(self):
  results["objective_values"] = -1 * results["objective_values"]
  results["best_objective"] = -1 * results["best_objective"]
 
- pbar.set_description("Best score: %s" % results["best_objective"])
+ if self.config.log_progress:
+ tqdm.write("Best score: %s" % results["best_objective"])
 
  # check if early stop criteria has been met
  if self.config.early_stop(results):
@@ -307,8 +313,12 @@ def runRandomOptimizer(self):
  random_hyper_parameters = self.ds.get_random_sample(n_iterations * batch_size)
 
  # running the iterations
- pbar = tqdm(range(0, len(random_hyper_parameters), batch_size))
- for idx in pbar:
+ iterator = (
+ tqdm(range(0, len(random_hyper_parameters), batch_size))
+ if self.config.log_progress
+ else range(0, len(random_hyper_parameters), batch_size)
+ )
+ for idx in iterator:
  # getting batch by batch random values to try
  batch_hyper_parameters = random_hyper_parameters[idx : idx + batch_size]
  X_list, Y_list = self.runUserObjective(batch_hyper_parameters)
@@ -327,7 +337,8 @@ def runRandomOptimizer(self):
  results["objective_values"] = -1 * results["objective_values"]
  results["best_objective"] = -1 * results["best_objective"]
 
- pbar.set_description("Best score: %s" % results["best_objective"])
+ if self.config.log_progress:
+ tqdm.write("Best score: %s" % results["best_objective"])
 
  # check if early stop criteria has been met
  if self.config.early_stop(results):

diff --git a/tests/test_tuner.py b/tests/test_tuner.py
@@ -6,6 +6,7 @@
 
 from mango.domain.domain_space import DomainSpace
 from mango import Tuner, scheduler
+import mango.tuner as tuner_mod
 from scipy.stats import uniform, dirichlet
 
 # Simple param_dict
@@ -156,6 +157,55 @@ def check(results, error_msg):
  check(results, "error while minimizing random")
 
 
+def test_log_progress_flag(monkeypatch):
+ """log_progress controls use of tqdm in the optimizer."""
+
+ class DummyTqdm:
+ def __init__(self):
+ self.iter_calls = 0
+ self.write_calls = 0
+
+ def __call__(self, iterable, *args, **kwargs):
+ # mimic tqdm(iterable) by returning the iterable unchanged
+ self.iter_calls += 1
+ return iterable
+
+ def write(self, *args, **kwargs):
+ self.write_calls += 1
+
+ dummy_tqdm = DummyTqdm()
+ monkeypatch.setattr(tuner_mod, "tqdm", dummy_tqdm)
+
+ param_dict = dict(x=range(-2, 3))
+
+ def objfunc(p_list):
+ # trivial, fast objective
+ return [0.0 for _ in p_list]
+
+ # When log_progress is False, tqdm should not be used
+ tuner = Tuner(
+ param_dict,
+ objfunc,
+ conf_dict=dict(num_iteration=3, initial_random=1, log_progress=False),
+ )
+ tuner.run()
+ assert dummy_tqdm.iter_calls == 0
+ assert dummy_tqdm.write_calls == 0
+
+ # When log_progress is True, tqdm should be used
+ dummy_tqdm.iter_calls = 0
+ dummy_tqdm.write_calls = 0
+
+ tuner = Tuner(
+ param_dict,
+ objfunc,
+ conf_dict=dict(num_iteration=3, initial_random=1, log_progress=True),
+ )
+ tuner.run()
+ assert dummy_tqdm.iter_calls > 0
+ assert dummy_tqdm.write_calls > 0
+
+
 def test_convex():
  param_dict = {
  "x": range(-100, 10),

diff --git a/README.md b/README.md
@@ -396,6 +396,7 @@ Note that if `initial_custom` option is given then `initial_random` is ignored.
 }
 ```
 However, use this option with caution as it could have unintended consequences. 
+- *log_progress*: `True` or `False` (default: `True`). When `True`, Mango logs optimization progress with a tqdm-based progress bar and per-iteration best scores. Set this to `False` to suppress progress logging, which is useful in CI environments or when you want cleaner logs.
 
 
 

diff --git a/development.md b/development.md
@@ -7,4 +7,24 @@ poetry config pypi-token.pypi <api_token>
 > poetry version [major, minor, patch]
 > poetry build
 > poetry publish
+```
+
+### Local development: virtualenv and tests
+
+Create a virtual environment in the project root:
+```shell
+python3 -m venv venv
+```
+
+Activate it from **fish**:
+```shell
+source ./venv/bin/activate.fish
+```
+
+Install development dependencies and run the test suite:
+```shell
+pip install --upgrade pip
+pip install poetry
+poetry install
+pytest
 ```
diff --git a/examples/__init__.py b/examples/__init__.py
diff --git a/examples/classifiers/__init__.py b/examples/classifiers/__init__.py
diff --git a/mango/domain/__init__.py b/mango/domain/__init__.py
diff --git a/mango/metatuner.py b/mango/metatuner.py
@@ -22,6 +22,7 @@ class MetaTuner:
  class Config:
  n_iter: int = 20
  n_init: int = 2
+ log_progress: bool = True
 
  def __init__(self, param_dict_list, objective_list, **kwargs):
  self.param_dict_list = param_dict_list
@@ -30,6 +31,7 @@ def __init__(self, param_dict_list, objective_list, **kwargs):
  # adjustable parameters
  self.num_of_iterations = self.config.n_iter
  self.initial_random = self.config.n_init
+ self.log_progress = self.config.log_progress
 
  # list of GPR for each objective
  self.gpr_list = []
@@ -222,9 +224,13 @@ def runExponentialTuner(self):
  # print(Optimizer_exploration)
 
  # Now run the optimization iterations
- pbar = tqdm(range(self.num_of_iterations))
+ iterator = (
+ tqdm(range(self.num_of_iterations))
+ if self.log_progress
+ else range(self.num_of_iterations)
+ )
 
- for itr in pbar:
+ for itr in iterator:
  # next values of x returned from individual function
  # Values in x are dependent on types of param dict, so using a list
  x_values_list = []
@@ -375,7 +381,8 @@ def runExponentialTuner(self):
  Optimizer_exploration[i] + 1.1 * self.exploration_rate
  )
 
- pbar.set_description(": Best score: %s" % max_val_y)
+ if self.log_progress:
+ tqdm.write(": Best score: %s" % max_val_y)
 
  # print(itr, s_values_list, Optimizer_iteration, Optimizer_exploration, max_val_y)#, Y_dict_array[selected_obj])
 

diff --git a/mango/optimizer/__init__.py b/mango/optimizer/__init__.py
diff --git a/mango/tuner.py b/mango/tuner.py
@@ -45,6 +45,7 @@ class TunerConfig:
  constraint: Callable = None
  param_sampler: Callable = parameter_sampler
  scale_params: bool = False
+ log_progress: bool = True
 
  def __post_init__(self):
  if self.optimizer not in self.valid_optimizers:
@@ -202,8 +203,12 @@ def runBayesianOptimizer(self):
  X_domain_np = self.ds.convert_GP_space(domain_list)
 
  # running the iterations
- pbar = tqdm(range(self.config.num_iteration))
- for _ in pbar:
+ iterator = (
+ tqdm(range(self.config.num_iteration))
+ if self.config.log_progress
+ else range(self.config.num_iteration)
+ )
+ for _ in iterator:
 
  # adding a Minimum exploration to explore independent of UCB
  if random.random() < self.config.exploration:
@@ -284,7 +289,8 @@ def runBayesianOptimizer(self):
  results["objective_values"] = -1 * results["objective_values"]
  results["best_objective"] = -1 * results["best_objective"]
 
- pbar.set_description("Best score: %s" % results["best_objective"])
+ if self.config.log_progress:
+ tqdm.write("Best score: %s" % results["best_objective"])
 
  # check if early stop criteria has been met
  if self.config.early_stop(results):
@@ -307,8 +313,12 @@ def runRandomOptimizer(self):
  random_hyper_parameters = self.ds.get_random_sample(n_iterations * batch_size)
 
  # running the iterations
- pbar = tqdm(range(0, len(random_hyper_parameters), batch_size))
- for idx in pbar:
+ iterator = (
+ tqdm(range(0, len(random_hyper_parameters), batch_size))
+ if self.config.log_progress
+ else range(0, len(random_hyper_parameters), batch_size)
+ )
+ for idx in iterator:
  # getting batch by batch random values to try
  batch_hyper_parameters = random_hyper_parameters[idx : idx + batch_size]
  X_list, Y_list = self.runUserObjective(batch_hyper_parameters)
@@ -327,7 +337,8 @@ def runRandomOptimizer(self):
  results["objective_values"] = -1 * results["objective_values"]
  results["best_objective"] = -1 * results["best_objective"]
 
- pbar.set_description("Best score: %s" % results["best_objective"])
+ if self.config.log_progress:
+ tqdm.write("Best score: %s" % results["best_objective"])
 
  # check if early stop criteria has been met
  if self.config.early_stop(results):

diff --git a/tests/test_tuner.py b/tests/test_tuner.py
@@ -6,6 +6,7 @@
 
 from mango.domain.domain_space import DomainSpace
 from mango import Tuner, scheduler
+import mango.tuner as tuner_mod
 from scipy.stats import uniform, dirichlet
 
 # Simple param_dict
@@ -156,6 +157,55 @@ def check(results, error_msg):
  check(results, "error while minimizing random")
 
 
+def test_log_progress_flag(monkeypatch):
+ """log_progress controls use of tqdm in the optimizer."""
+
+ class DummyTqdm:
+ def __init__(self):
+ self.iter_calls = 0
+ self.write_calls = 0
+
+ def __call__(self, iterable, *args, **kwargs):
+ # mimic tqdm(iterable) by returning the iterable unchanged
+ self.iter_calls += 1
+ return iterable
+
+ def write(self, *args, **kwargs):
+ self.write_calls += 1
+
+ dummy_tqdm = DummyTqdm()
+ monkeypatch.setattr(tuner_mod, "tqdm", dummy_tqdm)
+
+ param_dict = dict(x=range(-2, 3))
+
+ def objfunc(p_list):
+ # trivial, fast objective
+ return [0.0 for _ in p_list]
+
+ # When log_progress is False, tqdm should not be used
+ tuner = Tuner(
+ param_dict,
+ objfunc,
+ conf_dict=dict(num_iteration=3, initial_random=1, log_progress=False),
+ )
+ tuner.run()
+ assert dummy_tqdm.iter_calls == 0
+ assert dummy_tqdm.write_calls == 0
+
+ # When log_progress is True, tqdm should be used
+ dummy_tqdm.iter_calls = 0
+ dummy_tqdm.write_calls = 0
+
+ tuner = Tuner(
+ param_dict,
+ objfunc,
+ conf_dict=dict(num_iteration=3, initial_random=1, log_progress=True),
+ )
+ tuner.run()
+ assert dummy_tqdm.iter_calls > 0
+ assert dummy_tqdm.write_calls > 0
+
+
 def test_convex():
  param_dict = {
  "x": range(-100, 10),

diff --git a/README.md b/README.md
@@ -396,6 +396,7 @@ Note that if `initial_custom` option is given then `initial_random` is ignored.
 }
 ```
 However, use this option with caution as it could have unintended consequences. 
+- *log_progress*: `True` or `False` (default: `True`). When `True`, Mango logs optimization progress with a tqdm-based progress bar and per-iteration best scores. Set this to `False` to suppress progress logging, which is useful in CI environments or when you want cleaner logs.
 
 
 

diff --git a/development.md b/development.md
@@ -7,4 +7,24 @@ poetry config pypi-token.pypi <api_token>
 > poetry version [major, minor, patch]
 > poetry build
 > poetry publish
+```
+
+### Local development: virtualenv and tests
+
+Create a virtual environment in the project root:
+```shell
+python3 -m venv venv
+```
+
+Activate it from **fish**:
+```shell
+source ./venv/bin/activate.fish
+```
+
+Install development dependencies and run the test suite:
+```shell
+pip install --upgrade pip
+pip install poetry
+poetry install
+pytest
 ```
diff --git a/examples/__init__.py b/examples/__init__.py
diff --git a/examples/classifiers/__init__.py b/examples/classifiers/__init__.py
diff --git a/mango/domain/__init__.py b/mango/domain/__init__.py
diff --git a/mango/metatuner.py b/mango/metatuner.py
@@ -22,6 +22,7 @@ class MetaTuner:
  class Config:
  n_iter: int = 20
  n_init: int = 2
+ log_progress: bool = True
 
  def __init__(self, param_dict_list, objective_list, **kwargs):
  self.param_dict_list = param_dict_list
@@ -30,6 +31,7 @@ def __init__(self, param_dict_list, objective_list, **kwargs):
  # adjustable parameters
  self.num_of_iterations = self.config.n_iter
  self.initial_random = self.config.n_init
+ self.log_progress = self.config.log_progress
 
  # list of GPR for each objective
  self.gpr_list = []
@@ -222,9 +224,13 @@ def runExponentialTuner(self):
  # print(Optimizer_exploration)
 
  # Now run the optimization iterations
- pbar = tqdm(range(self.num_of_iterations))
+ iterator = (
+ tqdm(range(self.num_of_iterations))
+ if self.log_progress
+ else range(self.num_of_iterations)
+ )
 
- for itr in pbar:
+ for itr in iterator:
  # next values of x returned from individual function
  # Values in x are dependent on types of param dict, so using a list
  x_values_list = []
@@ -375,7 +381,8 @@ def runExponentialTuner(self):
  Optimizer_exploration[i] + 1.1 * self.exploration_rate
  )
 
- pbar.set_description(": Best score: %s" % max_val_y)
+ if self.log_progress:
+ tqdm.write(": Best score: %s" % max_val_y)
 
  # print(itr, s_values_list, Optimizer_iteration, Optimizer_exploration, max_val_y)#, Y_dict_array[selected_obj])
 

diff --git a/mango/optimizer/__init__.py b/mango/optimizer/__init__.py
diff --git a/mango/tuner.py b/mango/tuner.py
@@ -45,6 +45,7 @@ class TunerConfig:
  constraint: Callable = None
  param_sampler: Callable = parameter_sampler
  scale_params: bool = False
+ log_progress: bool = True
 
  def __post_init__(self):
  if self.optimizer not in self.valid_optimizers:
@@ -202,8 +203,12 @@ def runBayesianOptimizer(self):
  X_domain_np = self.ds.convert_GP_space(domain_list)
 
  # running the iterations
- pbar = tqdm(range(self.config.num_iteration))
- for _ in pbar:
+ iterator = (
+ tqdm(range(self.config.num_iteration))
+ if self.config.log_progress
+ else range(self.config.num_iteration)
+ )
+ for _ in iterator:
 
  # adding a Minimum exploration to explore independent of UCB
  if random.random() < self.config.exploration:
@@ -284,7 +289,8 @@ def runBayesianOptimizer(self):
  results["objective_values"] = -1 * results["objective_values"]
  results["best_objective"] = -1 * results["best_objective"]
 
- pbar.set_description("Best score: %s" % results["best_objective"])
+ if self.config.log_progress:
+ tqdm.write("Best score: %s" % results["best_objective"])
 
  # check if early stop criteria has been met
  if self.config.early_stop(results):
@@ -307,8 +313,12 @@ def runRandomOptimizer(self):
  random_hyper_parameters = self.ds.get_random_sample(n_iterations * batch_size)
 
  # running the iterations
- pbar = tqdm(range(0, len(random_hyper_parameters), batch_size))
- for idx in pbar:
+ iterator = (
+ tqdm(range(0, len(random_hyper_parameters), batch_size))
+ if self.config.log_progress
+ else range(0, len(random_hyper_parameters), batch_size)
+ )
+ for idx in iterator:
  # getting batch by batch random values to try
  batch_hyper_parameters = random_hyper_parameters[idx : idx + batch_size]
  X_list, Y_list = self.runUserObjective(batch_hyper_parameters)
@@ -327,7 +337,8 @@ def runRandomOptimizer(self):
  results["objective_values"] = -1 * results["objective_values"]
  results["best_objective"] = -1 * results["best_objective"]
 
- pbar.set_description("Best score: %s" % results["best_objective"])
+ if self.config.log_progress:
+ tqdm.write("Best score: %s" % results["best_objective"])
 
  # check if early stop criteria has been met
  if self.config.early_stop(results):

diff --git a/tests/test_tuner.py b/tests/test_tuner.py
@@ -6,6 +6,7 @@
 
 from mango.domain.domain_space import DomainSpace
 from mango import Tuner, scheduler
+import mango.tuner as tuner_mod
 from scipy.stats import uniform, dirichlet
 
 # Simple param_dict
@@ -156,6 +157,55 @@ def check(results, error_msg):
  check(results, "error while minimizing random")
 
 
+def test_log_progress_flag(monkeypatch):
+ """log_progress controls use of tqdm in the optimizer."""
+
+ class DummyTqdm:
+ def __init__(self):
+ self.iter_calls = 0
+ self.write_calls = 0
+
+ def __call__(self, iterable, *args, **kwargs):
+ # mimic tqdm(iterable) by returning the iterable unchanged
+ self.iter_calls += 1
+ return iterable
+
+ def write(self, *args, **kwargs):
+ self.write_calls += 1
+
+ dummy_tqdm = DummyTqdm()
+ monkeypatch.setattr(tuner_mod, "tqdm", dummy_tqdm)
+
+ param_dict = dict(x=range(-2, 3))
+
+ def objfunc(p_list):
+ # trivial, fast objective
+ return [0.0 for _ in p_list]
+
+ # When log_progress is False, tqdm should not be used
+ tuner = Tuner(
+ param_dict,
+ objfunc,
+ conf_dict=dict(num_iteration=3, initial_random=1, log_progress=False),
+ )
+ tuner.run()
+ assert dummy_tqdm.iter_calls == 0
+ assert dummy_tqdm.write_calls == 0
+
+ # When log_progress is True, tqdm should be used
+ dummy_tqdm.iter_calls = 0
+ dummy_tqdm.write_calls = 0
+
+ tuner = Tuner(
+ param_dict,
+ objfunc,
+ conf_dict=dict(num_iteration=3, initial_random=1, log_progress=True),
+ )
+ tuner.run()
+ assert dummy_tqdm.iter_calls > 0
+ assert dummy_tqdm.write_calls > 0
+
+
 def test_convex():
  param_dict = {
  "x": range(-100, 10),

diff --git a/README.md b/README.md
@@ -396,6 +396,7 @@ Note that if `initial_custom` option is given then `initial_random` is ignored.
 }
 ```
 However, use this option with caution as it could have unintended consequences. 
+- *log_progress*: `True` or `False` (default: `True`). When `True`, Mango logs optimization progress with a tqdm-based progress bar and per-iteration best scores. Set this to `False` to suppress progress logging, which is useful in CI environments or when you want cleaner logs.
 
 
 

diff --git a/development.md b/development.md
@@ -7,4 +7,24 @@ poetry config pypi-token.pypi <api_token>
 > poetry version [major, minor, patch]
 > poetry build
 > poetry publish
+```
+
+### Local development: virtualenv and tests
+
+Create a virtual environment in the project root:
+```shell
+python3 -m venv venv
+```
+
+Activate it from **fish**:
+```shell
+source ./venv/bin/activate.fish
+```
+
+Install development dependencies and run the test suite:
+```shell
+pip install --upgrade pip
+pip install poetry
+poetry install
+pytest
 ```
diff --git a/examples/__init__.py b/examples/__init__.py
diff --git a/examples/classifiers/__init__.py b/examples/classifiers/__init__.py
diff --git a/mango/domain/__init__.py b/mango/domain/__init__.py
diff --git a/mango/metatuner.py b/mango/metatuner.py
@@ -22,6 +22,7 @@ class MetaTuner:
  class Config:
  n_iter: int = 20
  n_init: int = 2
+ log_progress: bool = True
 
  def __init__(self, param_dict_list, objective_list, **kwargs):
  self.param_dict_list = param_dict_list
@@ -30,6 +31,7 @@ def __init__(self, param_dict_list, objective_list, **kwargs):
  # adjustable parameters
  self.num_of_iterations = self.config.n_iter
  self.initial_random = self.config.n_init
+ self.log_progress = self.config.log_progress
 
  # list of GPR for each objective
  self.gpr_list = []
@@ -222,9 +224,13 @@ def runExponentialTuner(self):
  # print(Optimizer_exploration)
 
  # Now run the optimization iterations
- pbar = tqdm(range(self.num_of_iterations))
+ iterator = (
+ tqdm(range(self.num_of_iterations))
+ if self.log_progress
+ else range(self.num_of_iterations)
+ )
 
- for itr in pbar:
+ for itr in iterator:
  # next values of x returned from individual function
  # Values in x are dependent on types of param dict, so using a list
  x_values_list = []
@@ -375,7 +381,8 @@ def runExponentialTuner(self):
  Optimizer_exploration[i] + 1.1 * self.exploration_rate
  )
 
- pbar.set_description(": Best score: %s" % max_val_y)
+ if self.log_progress:
+ tqdm.write(": Best score: %s" % max_val_y)
 
  # print(itr, s_values_list, Optimizer_iteration, Optimizer_exploration, max_val_y)#, Y_dict_array[selected_obj])
 

diff --git a/mango/optimizer/__init__.py b/mango/optimizer/__init__.py
diff --git a/mango/tuner.py b/mango/tuner.py
@@ -45,6 +45,7 @@ class TunerConfig:
  constraint: Callable = None
  param_sampler: Callable = parameter_sampler
  scale_params: bool = False
+ log_progress: bool = True
 
  def __post_init__(self):
  if self.optimizer not in self.valid_optimizers:
@@ -202,8 +203,12 @@ def runBayesianOptimizer(self):
  X_domain_np = self.ds.convert_GP_space(domain_list)
 
  # running the iterations
- pbar = tqdm(range(self.config.num_iteration))
- for _ in pbar:
+ iterator = (
+ tqdm(range(self.config.num_iteration))
+ if self.config.log_progress
+ else range(self.config.num_iteration)
+ )
+ for _ in iterator:
 
  # adding a Minimum exploration to explore independent of UCB
  if random.random() < self.config.exploration:
@@ -284,7 +289,8 @@ def runBayesianOptimizer(self):
  results["objective_values"] = -1 * results["objective_values"]
  results["best_objective"] = -1 * results["best_objective"]
 
- pbar.set_description("Best score: %s" % results["best_objective"])
+ if self.config.log_progress:
+ tqdm.write("Best score: %s" % results["best_objective"])
 
  # check if early stop criteria has been met
  if self.config.early_stop(results):
@@ -307,8 +313,12 @@ def runRandomOptimizer(self):
  random_hyper_parameters = self.ds.get_random_sample(n_iterations * batch_size)
 
  # running the iterations
- pbar = tqdm(range(0, len(random_hyper_parameters), batch_size))
- for idx in pbar:
+ iterator = (
+ tqdm(range(0, len(random_hyper_parameters), batch_size))
+ if self.config.log_progress
+ else range(0, len(random_hyper_parameters), batch_size)
+ )
+ for idx in iterator:
  # getting batch by batch random values to try
  batch_hyper_parameters = random_hyper_parameters[idx : idx + batch_size]
  X_list, Y_list = self.runUserObjective(batch_hyper_parameters)
@@ -327,7 +337,8 @@ def runRandomOptimizer(self):
  results["objective_values"] = -1 * results["objective_values"]
  results["best_objective"] = -1 * results["best_objective"]
 
- pbar.set_description("Best score: %s" % results["best_objective"])
+ if self.config.log_progress:
+ tqdm.write("Best score: %s" % results["best_objective"])
 
  # check if early stop criteria has been met
  if self.config.early_stop(results):

diff --git a/tests/test_tuner.py b/tests/test_tuner.py
@@ -6,6 +6,7 @@
 
 from mango.domain.domain_space import DomainSpace
 from mango import Tuner, scheduler
+import mango.tuner as tuner_mod
 from scipy.stats import uniform, dirichlet
 
 # Simple param_dict
@@ -156,6 +157,55 @@ def check(results, error_msg):
  check(results, "error while minimizing random")
 
 
+def test_log_progress_flag(monkeypatch):
+ """log_progress controls use of tqdm in the optimizer."""
+
+ class DummyTqdm:
+ def __init__(self):
+ self.iter_calls = 0
+ self.write_calls = 0
+
+ def __call__(self, iterable, *args, **kwargs):
+ # mimic tqdm(iterable) by returning the iterable unchanged
+ self.iter_calls += 1
+ return iterable
+
+ def write(self, *args, **kwargs):
+ self.write_calls += 1
+
+ dummy_tqdm = DummyTqdm()
+ monkeypatch.setattr(tuner_mod, "tqdm", dummy_tqdm)
+
+ param_dict = dict(x=range(-2, 3))
+
+ def objfunc(p_list):
+ # trivial, fast objective
+ return [0.0 for _ in p_list]
+
+ # When log_progress is False, tqdm should not be used
+ tuner = Tuner(
+ param_dict,
+ objfunc,
+ conf_dict=dict(num_iteration=3, initial_random=1, log_progress=False),
+ )
+ tuner.run()
+ assert dummy_tqdm.iter_calls == 0
+ assert dummy_tqdm.write_calls == 0
+
+ # When log_progress is True, tqdm should be used
+ dummy_tqdm.iter_calls = 0
+ dummy_tqdm.write_calls = 0
+
+ tuner = Tuner(
+ param_dict,
+ objfunc,
+ conf_dict=dict(num_iteration=3, initial_random=1, log_progress=True),
+ )
+ tuner.run()
+ assert dummy_tqdm.iter_calls > 0
+ assert dummy_tqdm.write_calls > 0
+
+
 def test_convex():
  param_dict = {
  "x": range(-100, 10),

diff --git a/README.md b/README.md
@@ -396,6 +396,7 @@ Note that if `initial_custom` option is given then `initial_random` is ignored.
 }
 ```
 However, use this option with caution as it could have unintended consequences. 
+- *log_progress*: `True` or `False` (default: `True`). When `True`, Mango logs optimization progress with a tqdm-based progress bar and per-iteration best scores. Set this to `False` to suppress progress logging, which is useful in CI environments or when you want cleaner logs.
 
 
 

diff --git a/development.md b/development.md
@@ -7,4 +7,24 @@ poetry config pypi-token.pypi <api_token>
 > poetry version [major, minor, patch]
 > poetry build
 > poetry publish
+```
+
+### Local development: virtualenv and tests
+
+Create a virtual environment in the project root:
+```shell
+python3 -m venv venv
+```
+
+Activate it from **fish**:
+```shell
+source ./venv/bin/activate.fish
+```
+
+Install development dependencies and run the test suite:
+```shell
+pip install --upgrade pip
+pip install poetry
+poetry install
+pytest
 ```
diff --git a/examples/__init__.py b/examples/__init__.py
diff --git a/examples/classifiers/__init__.py b/examples/classifiers/__init__.py
diff --git a/mango/domain/__init__.py b/mango/domain/__init__.py
diff --git a/mango/metatuner.py b/mango/metatuner.py
@@ -22,6 +22,7 @@ class MetaTuner:
  class Config:
  n_iter: int = 20
  n_init: int = 2
+ log_progress: bool = True
 
  def __init__(self, param_dict_list, objective_list, **kwargs):
  self.param_dict_list = param_dict_list
@@ -30,6 +31,7 @@ def __init__(self, param_dict_list, objective_list, **kwargs):
  # adjustable parameters
  self.num_of_iterations = self.config.n_iter
  self.initial_random = self.config.n_init
+ self.log_progress = self.config.log_progress
 
  # list of GPR for each objective
  self.gpr_list = []
@@ -222,9 +224,13 @@ def runExponentialTuner(self):
  # print(Optimizer_exploration)
 
  # Now run the optimization iterations
- pbar = tqdm(range(self.num_of_iterations))
+ iterator = (
+ tqdm(range(self.num_of_iterations))
+ if self.log_progress
+ else range(self.num_of_iterations)
+ )
 
- for itr in pbar:
+ for itr in iterator:
  # next values of x returned from individual function
  # Values in x are dependent on types of param dict, so using a list
  x_values_list = []
@@ -375,7 +381,8 @@ def runExponentialTuner(self):
  Optimizer_exploration[i] + 1.1 * self.exploration_rate
  )
 
- pbar.set_description(": Best score: %s" % max_val_y)
+ if self.log_progress:
+ tqdm.write(": Best score: %s" % max_val_y)
 
  # print(itr, s_values_list, Optimizer_iteration, Optimizer_exploration, max_val_y)#, Y_dict_array[selected_obj])
 

diff --git a/mango/optimizer/__init__.py b/mango/optimizer/__init__.py
diff --git a/mango/tuner.py b/mango/tuner.py
@@ -45,6 +45,7 @@ class TunerConfig:
  constraint: Callable = None
  param_sampler: Callable = parameter_sampler
  scale_params: bool = False
+ log_progress: bool = True
 
  def __post_init__(self):
  if self.optimizer not in self.valid_optimizers:
@@ -202,8 +203,12 @@ def runBayesianOptimizer(self):
  X_domain_np = self.ds.convert_GP_space(domain_list)
 
  # running the iterations
- pbar = tqdm(range(self.config.num_iteration))
- for _ in pbar:
+ iterator = (
+ tqdm(range(self.config.num_iteration))
+ if self.config.log_progress
+ else range(self.config.num_iteration)
+ )
+ for _ in iterator:
 
  # adding a Minimum exploration to explore independent of UCB
  if random.random() < self.config.exploration:
@@ -284,7 +289,8 @@ def runBayesianOptimizer(self):
  results["objective_values"] = -1 * results["objective_values"]
  results["best_objective"] = -1 * results["best_objective"]
 
- pbar.set_description("Best score: %s" % results["best_objective"])
+ if self.config.log_progress:
+ tqdm.write("Best score: %s" % results["best_objective"])
 
  # check if early stop criteria has been met
  if self.config.early_stop(results):
@@ -307,8 +313,12 @@ def runRandomOptimizer(self):
  random_hyper_parameters = self.ds.get_random_sample(n_iterations * batch_size)
 
  # running the iterations
- pbar = tqdm(range(0, len(random_hyper_parameters), batch_size))
- for idx in pbar:
+ iterator = (
+ tqdm(range(0, len(random_hyper_parameters), batch_size))
+ if self.config.log_progress
+ else range(0, len(random_hyper_parameters), batch_size)
+ )
+ for idx in iterator:
  # getting batch by batch random values to try
  batch_hyper_parameters = random_hyper_parameters[idx : idx + batch_size]
  X_list, Y_list = self.runUserObjective(batch_hyper_parameters)
@@ -327,7 +337,8 @@ def runRandomOptimizer(self):
  results["objective_values"] = -1 * results["objective_values"]
  results["best_objective"] = -1 * results["best_objective"]
 
- pbar.set_description("Best score: %s" % results["best_objective"])
+ if self.config.log_progress:
+ tqdm.write("Best score: %s" % results["best_objective"])
 
  # check if early stop criteria has been met
  if self.config.early_stop(results):

diff --git a/tests/test_tuner.py b/tests/test_tuner.py
@@ -6,6 +6,7 @@
 
 from mango.domain.domain_space import DomainSpace
 from mango import Tuner, scheduler
+import mango.tuner as tuner_mod
 from scipy.stats import uniform, dirichlet
 
 # Simple param_dict
@@ -156,6 +157,55 @@ def check(results, error_msg):
  check(results, "error while minimizing random")
 
 
+def test_log_progress_flag(monkeypatch):
+ """log_progress controls use of tqdm in the optimizer."""
+
+ class DummyTqdm:
+ def __init__(self):
+ self.iter_calls = 0
+ self.write_calls = 0
+
+ def __call__(self, iterable, *args, **kwargs):
+ # mimic tqdm(iterable) by returning the iterable unchanged
+ self.iter_calls += 1
+ return iterable
+
+ def write(self, *args, **kwargs):
+ self.write_calls += 1
+
+ dummy_tqdm = DummyTqdm()
+ monkeypatch.setattr(tuner_mod, "tqdm", dummy_tqdm)
+
+ param_dict = dict(x=range(-2, 3))
+
+ def objfunc(p_list):
+ # trivial, fast objective
+ return [0.0 for _ in p_list]
+
+ # When log_progress is False, tqdm should not be used
+ tuner = Tuner(
+ param_dict,
+ objfunc,
+ conf_dict=dict(num_iteration=3, initial_random=1, log_progress=False),
+ )
+ tuner.run()
+ assert dummy_tqdm.iter_calls == 0
+ assert dummy_tqdm.write_calls == 0
+
+ # When log_progress is True, tqdm should be used
+ dummy_tqdm.iter_calls = 0
+ dummy_tqdm.write_calls = 0
+
+ tuner = Tuner(
+ param_dict,
+ objfunc,
+ conf_dict=dict(num_iteration=3, initial_random=1, log_progress=True),
+ )
+ tuner.run()
+ assert dummy_tqdm.iter_calls > 0
+ assert dummy_tqdm.write_calls > 0
+
+
 def test_convex():
  param_dict = {
  "x": range(-100, 10),

diff --git a/README.md b/README.md
@@ -396,6 +396,7 @@ Note that if `initial_custom` option is given then `initial_random` is ignored.
 }
 ```
 However, use this option with caution as it could have unintended consequences. 
+- *log_progress*: `True` or `False` (default: `True`). When `True`, Mango logs optimization progress with a tqdm-based progress bar and per-iteration best scores. Set this to `False` to suppress progress logging, which is useful in CI environments or when you want cleaner logs.
 
 
 

diff --git a/development.md b/development.md
@@ -7,4 +7,24 @@ poetry config pypi-token.pypi <api_token>
 > poetry version [major, minor, patch]
 > poetry build
 > poetry publish
+```
+
+### Local development: virtualenv and tests
+
+Create a virtual environment in the project root:
+```shell
+python3 -m venv venv
+```
+
+Activate it from **fish**:
+```shell
+source ./venv/bin/activate.fish
+```
+
+Install development dependencies and run the test suite:
+```shell
+pip install --upgrade pip
+pip install poetry
+poetry install
+pytest
 ```
diff --git a/examples/__init__.py b/examples/__init__.py
diff --git a/examples/classifiers/__init__.py b/examples/classifiers/__init__.py
diff --git a/mango/domain/__init__.py b/mango/domain/__init__.py
diff --git a/mango/metatuner.py b/mango/metatuner.py
@@ -22,6 +22,7 @@ class MetaTuner:
  class Config:
  n_iter: int = 20
  n_init: int = 2
+ log_progress: bool = True
 
  def __init__(self, param_dict_list, objective_list, **kwargs):
  self.param_dict_list = param_dict_list
@@ -30,6 +31,7 @@ def __init__(self, param_dict_list, objective_list, **kwargs):
  # adjustable parameters
  self.num_of_iterations = self.config.n_iter
  self.initial_random = self.config.n_init
+ self.log_progress = self.config.log_progress
 
  # list of GPR for each objective
  self.gpr_list = []
@@ -222,9 +224,13 @@ def runExponentialTuner(self):
  # print(Optimizer_exploration)
 
  # Now run the optimization iterations
- pbar = tqdm(range(self.num_of_iterations))
+ iterator = (
+ tqdm(range(self.num_of_iterations))
+ if self.log_progress
+ else range(self.num_of_iterations)
+ )
 
- for itr in pbar:
+ for itr in iterator:
  # next values of x returned from individual function
  # Values in x are dependent on types of param dict, so using a list
  x_values_list = []
@@ -375,7 +381,8 @@ def runExponentialTuner(self):
  Optimizer_exploration[i] + 1.1 * self.exploration_rate
  )
 
- pbar.set_description(": Best score: %s" % max_val_y)
+ if self.log_progress:
+ tqdm.write(": Best score: %s" % max_val_y)
 
  # print(itr, s_values_list, Optimizer_iteration, Optimizer_exploration, max_val_y)#, Y_dict_array[selected_obj])
 

diff --git a/mango/optimizer/__init__.py b/mango/optimizer/__init__.py
diff --git a/mango/tuner.py b/mango/tuner.py
@@ -45,6 +45,7 @@ class TunerConfig:
  constraint: Callable = None
  param_sampler: Callable = parameter_sampler
  scale_params: bool = False
+ log_progress: bool = True
 
  def __post_init__(self):
  if self.optimizer not in self.valid_optimizers:
@@ -202,8 +203,12 @@ def runBayesianOptimizer(self):
  X_domain_np = self.ds.convert_GP_space(domain_list)
 
  # running the iterations
- pbar = tqdm(range(self.config.num_iteration))
- for _ in pbar:
+ iterator = (
+ tqdm(range(self.config.num_iteration))
+ if self.config.log_progress
+ else range(self.config.num_iteration)
+ )
+ for _ in iterator:
 
  # adding a Minimum exploration to explore independent of UCB
  if random.random() < self.config.exploration:
@@ -284,7 +289,8 @@ def runBayesianOptimizer(self):
  results["objective_values"] = -1 * results["objective_values"]
  results["best_objective"] = -1 * results["best_objective"]
 
- pbar.set_description("Best score: %s" % results["best_objective"])
+ if self.config.log_progress:
+ tqdm.write("Best score: %s" % results["best_objective"])
 
  # check if early stop criteria has been met
  if self.config.early_stop(results):
@@ -307,8 +313,12 @@ def runRandomOptimizer(self):
  random_hyper_parameters = self.ds.get_random_sample(n_iterations * batch_size)
 
  # running the iterations
- pbar = tqdm(range(0, len(random_hyper_parameters), batch_size))
- for idx in pbar:
+ iterator = (
+ tqdm(range(0, len(random_hyper_parameters), batch_size))
+ if self.config.log_progress
+ else range(0, len(random_hyper_parameters), batch_size)
+ )
+ for idx in iterator:
  # getting batch by batch random values to try
  batch_hyper_parameters = random_hyper_parameters[idx : idx + batch_size]
  X_list, Y_list = self.runUserObjective(batch_hyper_parameters)
@@ -327,7 +337,8 @@ def runRandomOptimizer(self):
  results["objective_values"] = -1 * results["objective_values"]
  results["best_objective"] = -1 * results["best_objective"]
 
- pbar.set_description("Best score: %s" % results["best_objective"])
+ if self.config.log_progress:
+ tqdm.write("Best score: %s" % results["best_objective"])
 
  # check if early stop criteria has been met
  if self.config.early_stop(results):

diff --git a/tests/test_tuner.py b/tests/test_tuner.py
@@ -6,6 +6,7 @@
 
 from mango.domain.domain_space import DomainSpace
 from mango import Tuner, scheduler
+import mango.tuner as tuner_mod
 from scipy.stats import uniform, dirichlet
 
 # Simple param_dict
@@ -156,6 +157,55 @@ def check(results, error_msg):
  check(results, "error while minimizing random")
 
 
+def test_log_progress_flag(monkeypatch):
+ """log_progress controls use of tqdm in the optimizer."""
+
+ class DummyTqdm:
+ def __init__(self):
+ self.iter_calls = 0
+ self.write_calls = 0
+
+ def __call__(self, iterable, *args, **kwargs):
+ # mimic tqdm(iterable) by returning the iterable unchanged
+ self.iter_calls += 1
+ return iterable
+
+ def write(self, *args, **kwargs):
+ self.write_calls += 1
+
+ dummy_tqdm = DummyTqdm()
+ monkeypatch.setattr(tuner_mod, "tqdm", dummy_tqdm)
+
+ param_dict = dict(x=range(-2, 3))
+
+ def objfunc(p_list):
+ # trivial, fast objective
+ return [0.0 for _ in p_list]
+
+ # When log_progress is False, tqdm should not be used
+ tuner = Tuner(
+ param_dict,
+ objfunc,
+ conf_dict=dict(num_iteration=3, initial_random=1, log_progress=False),
+ )
+ tuner.run()
+ assert dummy_tqdm.iter_calls == 0
+ assert dummy_tqdm.write_calls == 0
+
+ # When log_progress is True, tqdm should be used
+ dummy_tqdm.iter_calls = 0
+ dummy_tqdm.write_calls = 0
+
+ tuner = Tuner(
+ param_dict,
+ objfunc,
+ conf_dict=dict(num_iteration=3, initial_random=1, log_progress=True),
+ )
+ tuner.run()
+ assert dummy_tqdm.iter_calls > 0
+ assert dummy_tqdm.write_calls > 0
+
+
 def test_convex():
  param_dict = {
  "x": range(-100, 10),

diff --git a/README.md b/README.md
@@ -396,6 +396,7 @@ Note that if `initial_custom` option is given then `initial_random` is ignored.
 }
 ```
 However, use this option with caution as it could have unintended consequences. 
+- *log_progress*: `True` or `False` (default: `True`). When `True`, Mango logs optimization progress with a tqdm-based progress bar and per-iteration best scores. Set this to `False` to suppress progress logging, which is useful in CI environments or when you want cleaner logs.
 
 
 

diff --git a/development.md b/development.md
@@ -7,4 +7,24 @@ poetry config pypi-token.pypi <api_token>
 > poetry version [major, minor, patch]
 > poetry build
 > poetry publish
+```
+
+### Local development: virtualenv and tests
+
+Create a virtual environment in the project root:
+```shell
+python3 -m venv venv
+```
+
+Activate it from **fish**:
+```shell
+source ./venv/bin/activate.fish
+```
+
+Install development dependencies and run the test suite:
+```shell
+pip install --upgrade pip
+pip install poetry
+poetry install
+pytest
 ```
diff --git a/examples/__init__.py b/examples/__init__.py
diff --git a/examples/classifiers/__init__.py b/examples/classifiers/__init__.py
diff --git a/mango/domain/__init__.py b/mango/domain/__init__.py
diff --git a/mango/metatuner.py b/mango/metatuner.py
@@ -22,6 +22,7 @@ class MetaTuner:
  class Config:
  n_iter: int = 20
  n_init: int = 2
+ log_progress: bool = True
 
  def __init__(self, param_dict_list, objective_list, **kwargs):
  self.param_dict_list = param_dict_list
@@ -30,6 +31,7 @@ def __init__(self, param_dict_list, objective_list, **kwargs):
  # adjustable parameters
  self.num_of_iterations = self.config.n_iter
  self.initial_random = self.config.n_init
+ self.log_progress = self.config.log_progress
 
  # list of GPR for each objective
  self.gpr_list = []
@@ -222,9 +224,13 @@ def runExponentialTuner(self):
  # print(Optimizer_exploration)
 
  # Now run the optimization iterations
- pbar = tqdm(range(self.num_of_iterations))
+ iterator = (
+ tqdm(range(self.num_of_iterations))
+ if self.log_progress
+ else range(self.num_of_iterations)
+ )
 
- for itr in pbar:
+ for itr in iterator:
  # next values of x returned from individual function
  # Values in x are dependent on types of param dict, so using a list
  x_values_list = []
@@ -375,7 +381,8 @@ def runExponentialTuner(self):
  Optimizer_exploration[i] + 1.1 * self.exploration_rate
  )
 
- pbar.set_description(": Best score: %s" % max_val_y)
+ if self.log_progress:
+ tqdm.write(": Best score: %s" % max_val_y)
 
  # print(itr, s_values_list, Optimizer_iteration, Optimizer_exploration, max_val_y)#, Y_dict_array[selected_obj])
 

diff --git a/mango/optimizer/__init__.py b/mango/optimizer/__init__.py
diff --git a/mango/tuner.py b/mango/tuner.py
@@ -45,6 +45,7 @@ class TunerConfig:
  constraint: Callable = None
  param_sampler: Callable = parameter_sampler
  scale_params: bool = False
+ log_progress: bool = True
 
  def __post_init__(self):
  if self.optimizer not in self.valid_optimizers:
@@ -202,8 +203,12 @@ def runBayesianOptimizer(self):
  X_domain_np = self.ds.convert_GP_space(domain_list)
 
  # running the iterations
- pbar = tqdm(range(self.config.num_iteration))
- for _ in pbar:
+ iterator = (
+ tqdm(range(self.config.num_iteration))
+ if self.config.log_progress
+ else range(self.config.num_iteration)
+ )
+ for _ in iterator:
 
  # adding a Minimum exploration to explore independent of UCB
  if random.random() < self.config.exploration:
@@ -284,7 +289,8 @@ def runBayesianOptimizer(self):
  results["objective_values"] = -1 * results["objective_values"]
  results["best_objective"] = -1 * results["best_objective"]
 
- pbar.set_description("Best score: %s" % results["best_objective"])
+ if self.config.log_progress:
+ tqdm.write("Best score: %s" % results["best_objective"])
 
  # check if early stop criteria has been met
  if self.config.early_stop(results):
@@ -307,8 +313,12 @@ def runRandomOptimizer(self):
  random_hyper_parameters = self.ds.get_random_sample(n_iterations * batch_size)
 
  # running the iterations
- pbar = tqdm(range(0, len(random_hyper_parameters), batch_size))
- for idx in pbar:
+ iterator = (
+ tqdm(range(0, len(random_hyper_parameters), batch_size))
+ if self.config.log_progress
+ else range(0, len(random_hyper_parameters), batch_size)
+ )
+ for idx in iterator:
  # getting batch by batch random values to try
  batch_hyper_parameters = random_hyper_parameters[idx : idx + batch_size]
  X_list, Y_list = self.runUserObjective(batch_hyper_parameters)
@@ -327,7 +337,8 @@ def runRandomOptimizer(self):
  results["objective_values"] = -1 * results["objective_values"]
  results["best_objective"] = -1 * results["best_objective"]
 
- pbar.set_description("Best score: %s" % results["best_objective"])
+ if self.config.log_progress:
+ tqdm.write("Best score: %s" % results["best_objective"])
 
  # check if early stop criteria has been met
  if self.config.early_stop(results):

diff --git a/tests/test_tuner.py b/tests/test_tuner.py
@@ -6,6 +6,7 @@
 
 from mango.domain.domain_space import DomainSpace
 from mango import Tuner, scheduler
+import mango.tuner as tuner_mod
 from scipy.stats import uniform, dirichlet
 
 # Simple param_dict
@@ -156,6 +157,55 @@ def check(results, error_msg):
  check(results, "error while minimizing random")
 
 
+def test_log_progress_flag(monkeypatch):
+ """log_progress controls use of tqdm in the optimizer."""
+
+ class DummyTqdm:
+ def __init__(self):
+ self.iter_calls = 0
+ self.write_calls = 0
+
+ def __call__(self, iterable, *args, **kwargs):
+ # mimic tqdm(iterable) by returning the iterable unchanged
+ self.iter_calls += 1
+ return iterable
+
+ def write(self, *args, **kwargs):
+ self.write_calls += 1
+
+ dummy_tqdm = DummyTqdm()
+ monkeypatch.setattr(tuner_mod, "tqdm", dummy_tqdm)
+
+ param_dict = dict(x=range(-2, 3))
+
+ def objfunc(p_list):
+ # trivial, fast objective
+ return [0.0 for _ in p_list]
+
+ # When log_progress is False, tqdm should not be used
+ tuner = Tuner(
+ param_dict,
+ objfunc,
+ conf_dict=dict(num_iteration=3, initial_random=1, log_progress=False),
+ )
+ tuner.run()
+ assert dummy_tqdm.iter_calls == 0
+ assert dummy_tqdm.write_calls == 0
+
+ # When log_progress is True, tqdm should be used
+ dummy_tqdm.iter_calls = 0
+ dummy_tqdm.write_calls = 0
+
+ tuner = Tuner(
+ param_dict,
+ objfunc,
+ conf_dict=dict(num_iteration=3, initial_random=1, log_progress=True),
+ )
+ tuner.run()
+ assert dummy_tqdm.iter_calls > 0
+ assert dummy_tqdm.write_calls > 0
+
+
 def test_convex():
  param_dict = {
  "x": range(-100, 10),

diff --git a/README.md b/README.md
@@ -396,6 +396,7 @@ Note that if `initial_custom` option is given then `initial_random` is ignored.
 }
 ```
 However, use this option with caution as it could have unintended consequences. 
+- *log_progress*: `True` or `False` (default: `True`). When `True`, Mango logs optimization progress with a tqdm-based progress bar and per-iteration best scores. Set this to `False` to suppress progress logging, which is useful in CI environments or when you want cleaner logs.
 
 
 

diff --git a/development.md b/development.md
@@ -7,4 +7,24 @@ poetry config pypi-token.pypi <api_token>
 > poetry version [major, minor, patch]
 > poetry build
 > poetry publish
+```
+
+### Local development: virtualenv and tests
+
+Create a virtual environment in the project root:
+```shell
+python3 -m venv venv
+```
+
+Activate it from **fish**:
+```shell
+source ./venv/bin/activate.fish
+```
+
+Install development dependencies and run the test suite:
+```shell
+pip install --upgrade pip
+pip install poetry
+poetry install
+pytest
 ```
diff --git a/examples/__init__.py b/examples/__init__.py
diff --git a/examples/classifiers/__init__.py b/examples/classifiers/__init__.py
diff --git a/mango/domain/__init__.py b/mango/domain/__init__.py
diff --git a/mango/metatuner.py b/mango/metatuner.py
@@ -22,6 +22,7 @@ class MetaTuner:
  class Config:
  n_iter: int = 20
  n_init: int = 2
+ log_progress: bool = True
 
  def __init__(self, param_dict_list, objective_list, **kwargs):
  self.param_dict_list = param_dict_list
@@ -30,6 +31,7 @@ def __init__(self, param_dict_list, objective_list, **kwargs):
  # adjustable parameters
  self.num_of_iterations = self.config.n_iter
  self.initial_random = self.config.n_init
+ self.log_progress = self.config.log_progress
 
  # list of GPR for each objective
  self.gpr_list = []
@@ -222,9 +224,13 @@ def runExponentialTuner(self):
  # print(Optimizer_exploration)
 
  # Now run the optimization iterations
- pbar = tqdm(range(self.num_of_iterations))
+ iterator = (
+ tqdm(range(self.num_of_iterations))
+ if self.log_progress
+ else range(self.num_of_iterations)
+ )
 
- for itr in pbar:
+ for itr in iterator:
  # next values of x returned from individual function
  # Values in x are dependent on types of param dict, so using a list
  x_values_list = []
@@ -375,7 +381,8 @@ def runExponentialTuner(self):
  Optimizer_exploration[i] + 1.1 * self.exploration_rate
  )
 
- pbar.set_description(": Best score: %s" % max_val_y)
+ if self.log_progress:
+ tqdm.write(": Best score: %s" % max_val_y)
 
  # print(itr, s_values_list, Optimizer_iteration, Optimizer_exploration, max_val_y)#, Y_dict_array[selected_obj])
 

diff --git a/mango/optimizer/__init__.py b/mango/optimizer/__init__.py
diff --git a/mango/tuner.py b/mango/tuner.py
@@ -45,6 +45,7 @@ class TunerConfig:
  constraint: Callable = None
  param_sampler: Callable = parameter_sampler
  scale_params: bool = False
+ log_progress: bool = True
 
  def __post_init__(self):
  if self.optimizer not in self.valid_optimizers:
@@ -202,8 +203,12 @@ def runBayesianOptimizer(self):
  X_domain_np = self.ds.convert_GP_space(domain_list)
 
  # running the iterations
- pbar = tqdm(range(self.config.num_iteration))
- for _ in pbar:
+ iterator = (
+ tqdm(range(self.config.num_iteration))
+ if self.config.log_progress
+ else range(self.config.num_iteration)
+ )
+ for _ in iterator:
 
  # adding a Minimum exploration to explore independent of UCB
  if random.random() < self.config.exploration:
@@ -284,7 +289,8 @@ def runBayesianOptimizer(self):
  results["objective_values"] = -1 * results["objective_values"]
  results["best_objective"] = -1 * results["best_objective"]
 
- pbar.set_description("Best score: %s" % results["best_objective"])
+ if self.config.log_progress:
+ tqdm.write("Best score: %s" % results["best_objective"])
 
  # check if early stop criteria has been met
  if self.config.early_stop(results):
@@ -307,8 +313,12 @@ def runRandomOptimizer(self):
  random_hyper_parameters = self.ds.get_random_sample(n_iterations * batch_size)
 
  # running the iterations
- pbar = tqdm(range(0, len(random_hyper_parameters), batch_size))
- for idx in pbar:
+ iterator = (
+ tqdm(range(0, len(random_hyper_parameters), batch_size))
+ if self.config.log_progress
+ else range(0, len(random_hyper_parameters), batch_size)
+ )
+ for idx in iterator:
  # getting batch by batch random values to try
  batch_hyper_parameters = random_hyper_parameters[idx : idx + batch_size]
  X_list, Y_list = self.runUserObjective(batch_hyper_parameters)
@@ -327,7 +337,8 @@ def runRandomOptimizer(self):
  results["objective_values"] = -1 * results["objective_values"]
  results["best_objective"] = -1 * results["best_objective"]
 
- pbar.set_description("Best score: %s" % results["best_objective"])
+ if self.config.log_progress:
+ tqdm.write("Best score: %s" % results["best_objective"])
 
  # check if early stop criteria has been met
  if self.config.early_stop(results):

diff --git a/tests/test_tuner.py b/tests/test_tuner.py
@@ -6,6 +6,7 @@
 
 from mango.domain.domain_space import DomainSpace
 from mango import Tuner, scheduler
+import mango.tuner as tuner_mod
 from scipy.stats import uniform, dirichlet
 
 # Simple param_dict
@@ -156,6 +157,55 @@ def check(results, error_msg):
  check(results, "error while minimizing random")
 
 
+def test_log_progress_flag(monkeypatch):
+ """log_progress controls use of tqdm in the optimizer."""
+
+ class DummyTqdm:
+ def __init__(self):
+ self.iter_calls = 0
+ self.write_calls = 0
+
+ def __call__(self, iterable, *args, **kwargs):
+ # mimic tqdm(iterable) by returning the iterable unchanged
+ self.iter_calls += 1
+ return iterable
+
+ def write(self, *args, **kwargs):
+ self.write_calls += 1
+
+ dummy_tqdm = DummyTqdm()
+ monkeypatch.setattr(tuner_mod, "tqdm", dummy_tqdm)
+
+ param_dict = dict(x=range(-2, 3))
+
+ def objfunc(p_list):
+ # trivial, fast objective
+ return [0.0 for _ in p_list]
+
+ # When log_progress is False, tqdm should not be used
+ tuner = Tuner(
+ param_dict,
+ objfunc,
+ conf_dict=dict(num_iteration=3, initial_random=1, log_progress=False),
+ )
+ tuner.run()
+ assert dummy_tqdm.iter_calls == 0
+ assert dummy_tqdm.write_calls == 0
+
+ # When log_progress is True, tqdm should be used
+ dummy_tqdm.iter_calls = 0
+ dummy_tqdm.write_calls = 0
+
+ tuner = Tuner(
+ param_dict,
+ objfunc,
+ conf_dict=dict(num_iteration=3, initial_random=1, log_progress=True),
+ )
+ tuner.run()
+ assert dummy_tqdm.iter_calls > 0
+ assert dummy_tqdm.write_calls > 0
+
+
 def test_convex():
  param_dict = {
  "x": range(-100, 10),
-Original file line number
+Diff line change
@@ Expand Up @@
 }
 ```
 However, use this option with caution as it could have unintended consequences.
+- *log_progress*: `True` or `False` (default: `True`). When `True`, Mango logs optimization progress with a tqdm-based progress bar and per-iteration best scores. Set this to `False` to suppress progress logging, which is useful in CI environments or when you want cleaner logs.
@@ Expand Down @@