The automl.search() keep crashing

[rachelchann]

The automl.search keep crushing after i have with error messages
ValueError: GradientBoostingQuantileRegressor(base_estimator=GradientBoostingRegressor(loss='quantile', n_estimators=30), quantiles=[0.16, 0.5, 0.84], random_state=209652396) has to be a regressor.
I wonder if it is my package installation problems or the code has problems.

My code is as follows:
`y_train = X_train["show"]
X_train = X_train.drop(["show", "index_a", "index_b"], axis=1)

y_test = X_test["show"]
X_test = X_test.drop(["show", "index_a", "index_b"], axis=1)

import evalml
from evalml import AutoMLSearch
from evalml import AutoMLSearch
from evalml.model_family import ModelFamily
from evalml.tuners import RandomSearchTuner
automl = AutoMLSearch(
X_train=X_train,
y_train=y_train,
problem_type="binary", # For binary classification
max_time=3600, # Search for 10 minutes
objective="f1", # Optimize for F1 score (you can choose others)
verbose=True, # Show progress in the console
n_jobs= -1,

)

Start the AutoML search

automl.search()the error message is as follows:Evaluating Baseline Pipeline: Mode Baseline Binary Classification Pipeline
Mode Baseline Binary Classification Pipeline:
Starting cross validation
Finished cross validation - mean F1: 0.000

ValueError Traceback (most recent call last)
Cell In[17], line 2
1 # Start the AutoML search
----> 2 automl.search()

File /opt/anaconda3/envs/leica/lib/python3.12/site-packages/evalml/automl/automl_search.py:1075, in AutoMLSearch.search(self, interactive_plot)
1073 try:
1074 if not loop_interrupted:
-> 1075 current_batch_pipelines = self.automl_algorithm.next_batch()
1076 except StopIteration:
1077 self.logger.info("AutoML Algorithm out of recommendations, ending")

File /opt/anaconda3/envs/leica/lib/python3.12/site-packages/evalml/automl/automl_algorithm/default_algorithm.py:466, in DefaultAlgorithm.next_batch(self)
464 else:
465 if self._batch_number == 0:
--> 466 next_batch = self._create_naive_pipelines()
467 elif self._batch_number == 1:
468 next_batch = self._create_naive_pipelines(use_features=True)

File /opt/anaconda3/envs/leica/lib/python3.12/site-packages/evalml/automl/automl_algorithm/default_algorithm.py:251, in DefaultAlgorithm._create_naive_pipelines(self, use_features)
232 pipelines = [
233 make_pipeline(
234 X=self.X,
(...) 248 for estimator in estimators
249 ]
250 pipelines = self._add_without_pipelines(pipelines, estimators, feature_selector)
--> 251 pipelines = self._init_pipelines_with_starter_params(pipelines)
252 return pipelines

File /opt/anaconda3/envs/leica/lib/python3.12/site-packages/evalml/automl/automl_algorithm/default_algorithm.py:206, in DefaultAlgorithm._init_pipelines_with_starter_params(self, pipelines)
204 next_batch = []
205 for pipeline in pipelines:
--> 206 self._create_tuner(pipeline)
207 starting_parameters = self._tuners[pipeline.name].get_starting_parameters(
208 self._hyperparameters,
209 self.random_seed,
210 )
211 parameters = self._transform_parameters(pipeline, starting_parameters)

File /opt/anaconda3/envs/leica/lib/python3.12/site-packages/evalml/automl/automl_algorithm/automl_algorithm.py:90, in AutoMLAlgorithm._create_tuner(self, pipeline)
86 """Creates a tuner given the input pipeline."""
87 pipeline_hyperparameters = pipeline.get_hyperparameter_ranges(
88 self._hyperparameters,
89 )
---> 90 self._tuners[pipeline.name] = self._tuner_class(
91 pipeline_hyperparameters,
92 random_seed=self.random_seed,
93 )

File /opt/anaconda3/envs/leica/lib/python3.12/site-packages/evalml/tuners/skopt_tuner.py:43, in SKOptTuner.init(self, pipeline_hyperparameter_ranges, random_seed)
41 def init(self, pipeline_hyperparameter_ranges, random_seed=0):
42 super().init(pipeline_hyperparameter_ranges, random_seed=random_seed)
---> 43 self.opt = Optimizer(
44 self._search_space_ranges,
45 "GBRT",
46 acq_optimizer="sampling",
47 random_state=random_seed,
48 )

File /opt/anaconda3/envs/leica/lib/python3.12/site-packages/skopt/optimizer/optimizer.py:257, in Optimizer.init(self, dimensions, base_estimator, n_random_starts, n_initial_points, initial_point_generator, n_jobs, acq_func, acq_optimizer, random_state, model_queue_size, space_constraint, acq_func_kwargs, acq_optimizer_kwargs, avoid_duplicates)
255 # check if regressor
256 if not is_regressor(base_estimator) and base_estimator is not None:
--> 257 raise ValueError("%s has to be a regressor." % base_estimator)
259 # treat per second acqusition function specially
260 is_multi_regressor = isinstance(base_estimator, MultiOutputRegressor)

ValueError: GradientBoostingQuantileRegressor(base_estimator=GradientBoostingRegressor(loss='quantile',
n_estimators=30),
quantiles=[0.16, 0.5, 0.84],
random_state=209652396) has to be a regressor.`