Is there another method of finding the best algorithm other than leaderboard

[asmgx]

Using leaderboard() has been giving me hard time and giving errors ( I created a ticket for them)

Is there any other way to find the best algorithm in my model?

P.S. using AutoSklearn version 0.14.6

[eddiebergman]

You can use show_models but it's likely to give you similar errors. You can use these three lines to get the models with their weights. But be a bit safer about it.

Fyi, the issue hasn't been fixed but it at least won't give an error. It's in development but I'm not sure how long until a release is made.

Best,
Eddie

[eddiebergman]

Here's the fixed version for reference if you need
https://github.com/automl/auto-sklearn/blob/development/autosklearn/estimators.py#L938

[asmgx]

Shall I update this file (estimators.py#L938 manually with the code you provided? will that fix the issue

[eddiebergman]

You can do that if you like, or, instead of calling def leaderboard() you can use the same logic as the code provided to get out what you need.

[asmgx]

still getting error

Traceback (most recent call last):
  File "<stdin>", line 2, in <module>
  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/estimators.py", line 764, in leaderboard
    for rkey, rval in self.automl_.runhistory_.data.items()
AttributeError: 'NoneType' object has no attribute 'runhistory_'
>>>

[eddiebergman]

You'll have to read and understand the code. This seems like is a general python error and not auto-sklearn specific. I'm not sure if you modified the actual estimators file but you don't need to

You can do something like:

# Your file where you are using Autosklearn
my_classifier = AutoSklearnClassifier(...)
my_classifier.fit(X, y)

runhistory = my_classifier.automl_.runhistory_

# ... do stuff

[eddiebergman]

Closing due to inactivity and the latest error is not autosklearn specific.

[asmgx]

Sorry for the delay, i have been trying to get the bottom of this

This method still does not show an algorithm that has been used.

This is the output I got from clf.automl_.runhistory_.data.items()

odict_items([
(RunKey(config_id=1, instance_id='{"task_id": "9a9e1fbd-c357-11ec-a24f-192ff4cebbef"}', seed=0, budget=6.25), RunValue(cost=1.0, time=32.70566391944885, status=<StatusType.CRASHED: 3>, starttime=1650754177.2785182, endtime=1650754210.0102727, additional_info={'traceback': 'Traceback (most recent call last):\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/evaluation/__init__.py", line 42, in fit_predict_try_except_decorator\n    return ta(queue=queue, **kwargs)\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/evaluation/train_evaluator.py", line 1408, in eval_iterative_cv\n    eval_cv(\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/evaluation/train_evaluator.py", line 1383, in eval_cv\n    evaluator.fit_predict_and_loss(iterative=iterative)\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/evaluation/train_evaluator.py", line 329, in fit_predict_and_loss\n    model.iterative_fit(\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/pipeline/base.py", line 127, in iterative_fit\n    self._final_estimator.iterative_fit(X, y, n_iter=n_iter,\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/pipeline/components/classification/__init__.py", line 149, in iterative_fit\n    return self.choice.iterative_fit(X, y, n_iter=n_iter, **fit_params)\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/pipeline/components/classification/gradient_boosting.py", line 125, in iterative_fit\n    self.estimator.fit(X, y, sample_weight=sample_weight)\n  File "/home/my/.local/lib/python3.8/site-packages/sklearn/ensemble/_hist_gradient_boosting/gradient_boosting.py", line 270, in fit\n    sample_weight_val) = train_test_split(\n  File "/home/my/.local/lib/python3.8/site-packages/sklearn/model_selection/_split.py", line 2197, in train_test_split\n    train, test = next(cv.split(X=arrays[0], y=stratify))\n  File "/home/my/.local/lib/python3.8/site-packages/sklearn/model_selection/_split.py", line 1387, in split\n    for train, test in self._iter_indices(X, y, groups):\n  File "/home/my/.local/lib/python3.8/site-packages/sklearn/model_selection/_split.py", line 1715, in _iter_indices\n    raise ValueError("The least populated class in y has only 1"\nValueError: The least populated class in y has only 1 member, which is too few. The minimum number of groups for any class cannot be less than 2.\n', 'error': "ValueError('The least populated class in y has only 1 member, which is too few. The minimum number of groups for any class cannot be less than 2.')", 'configuration_origin': 'Initial design'})), 
(RunKey(config_id=2, instance_id='{"task_id": "9a9e1fbd-c357-11ec-a24f-192ff4cebbef"}', seed=0, budget=6.25), RunValue(cost=0.0024999999999999914, time=59.02157926559448, status=<StatusType.DONOTADVANCE: 7>, starttime=1650754177.2943952, endtime=1650754237.3593912, additional_info={'duration': 53.925941705703735, 'num_run': 3, 'train_loss': 0.0, 'subprocess_stdout': 'The least populated class in y has only 2 members, which is less than n_splits=5.\nThe least populated class in y has only 2 members, which is less than n_splits=5.\nThe least populated class in y has only 2 members, which is less than n_splits=5.\nThe least populated class in y has only 2 members, which is less than n_splits=5.\nThe least populated class in y has only 2 members, which is less than n_splits=5.\n', 'info': 'Run stopped because of timeout.', 'learning_curve': [0.0024999999999999914, 0.0024999999999999914, 0.0024999999999999914], 'learning_curve_runtime': [43.16014623641968, 48.51548194885254, 53.925941705703735], 'train_learning_curve': [0.0, 0.00031250000000000445, 0.0], 'configuration_origin': 'Initial design'})), 
(RunKey(config_id=3, instance_id='{"task_id": "9a9e1fbd-c357-11ec-a24f-192ff4cebbef"}', seed=0, budget=6.25), RunValue(cost=1.0, time=59.029137134552, status=<StatusType.TIMEOUT: 2>, starttime=1650754177.3097827, endtime=1650754237.377532, additional_info={'error': 'Timeout', 'configuration_origin': 'Initial design'})), 
(RunKey(config_id=4, instance_id='{"task_id": "9a9e1fbd-c357-11ec-a24f-192ff4cebbef"}', seed=0, budget=6.25), RunValue(cost=0.013749999999999997, time=59.02944850921631, status=<StatusType.DONOTADVANCE: 7>, starttime=1650754177.332573, endtime=1650754237.6695342, additional_info={'duration': 53.8226203918457, 'num_run': 5, 'train_loss': 0.0015625000000000222, 'subprocess_stdout': 'The least populated class in y has only 2 members, which is less than n_splits=5.\nThe least populated class in y has only 2 members, which is less than n_splits=5.\nThe least populated class in y has only 2 members, which is less than n_splits=5.\nThe least populated class in y has only 2 members, which is less than n_splits=5.\n', 'info': 'Run stopped because of timeout.', 'learning_curve': [0.01874999999999998, 0.013749999999999997], 'learning_curve_runtime': [41.331947803497314, 53.8226203918457], 'train_learning_curve': [0.007812500000000023, 0.0015625000000000222], 'configuration_origin': 'Initial design'})), 
(RunKey(config_id=5, instance_id='{"task_id": "9a9e1fbd-c357-11ec-a24f-192ff4cebbef"}', seed=0, budget=6.25), RunValue(cost=1.0, time=23.6397864818573, status=<StatusType.CRASHED: 3>, starttime=1650754177.6396298, endtime=1650754201.3307548, additional_info={'traceback': 'Traceback (most recent call last):\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/evaluation/__init__.py", line 42, in fit_predict_try_except_decorator\n    return ta(queue=queue, **kwargs)\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/evaluation/train_evaluator.py", line 1408, in eval_iterative_cv\n    eval_cv(\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/evaluation/train_evaluator.py", line 1383, in eval_cv\n    evaluator.fit_predict_and_loss(iterative=iterative)\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/evaluation/train_evaluator.py", line 329, in fit_predict_and_loss\n    model.iterative_fit(\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/pipeline/base.py", line 127, in iterative_fit\n    self._final_estimator.iterative_fit(X, y, n_iter=n_iter,\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/pipeline/components/classification/__init__.py", line 149, in iterative_fit\n    return self.choice.iterative_fit(X, y, n_iter=n_iter, **fit_params)\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/pipeline/components/classification/mlp.py", line 144, in iterative_fit\n    self.estimator.fit(X, y)\n  File "/home/my/.local/lib/python3.8/site-packages/sklearn/neural_network/_multilayer_perceptron.py", line 673, in fit\n    return self._fit(X, y, incremental=False)\n  File "/home/my/.local/lib/python3.8/site-packages/sklearn/neural_network/_multilayer_perceptron.py", line 399, in _fit\n    self._fit_stochastic(X, y, activations, deltas, coef_grads,\n  File "/home/my/.local/lib/python3.8/site-packages/sklearn/neural_network/_multilayer_perceptron.py", line 525, in _fit_stochastic\n    X, X_val, y, y_val = train_test_split(\n  File "/home/my/.local/lib/python3.8/site-packages/sklearn/model_selection/_split.py", line 2197, in train_test_split\n    train, test = next(cv.split(X=arrays[0], y=stratify))\n  File "/home/my/.local/lib/python3.8/site-packages/sklearn/model_selection/_split.py", line 1387, in split\n    for train, test in self._iter_indices(X, y, groups):\n  File "/home/my/.local/lib/python3.8/site-packages/sklearn/model_selection/_split.py", line 1715, in _iter_indices\n    raise ValueError("The least populated class in y has only 1"\nValueError: The least populated class in y has only 1 member, which is too few. The minimum number of groups for any class cannot be less than 2.\n', 'error': "ValueError('The least populated class in y has only 1 member, which is too few. The minimum number of groups for any class cannot be less than 2.')", 'configuration_origin': 'Initial design'})), 
(RunKey(config_id=6, instance_id='{"task_id": "9a9e1fbd-c357-11ec-a24f-192ff4cebbef"}', seed=0, budget=6.25), RunValue(cost=0.010000000000000009, time=59.06597280502319, status=<StatusType.DONOTADVANCE: 7>, starttime=1650754177.6630983, endtime=1650754237.8171875, additional_info={'duration': 47.305580377578735, 'num_run': 7, 'train_loss': 0.005312500000000009, 'subprocess_stdout': 'The least populated class in y has only 2 members, which is less than n_splits=5.\nThe least populated class in y has only 2 members, which is less than n_splits=5.\nThe least populated class in y has only 2 members, which is less than n_splits=5.\n', 'info': 'Run stopped because of timeout.', 'configuration_origin': 'Initial design'})), 
(RunKey(config_id=7, instance_id='{"task_id": "9a9e1fbd-c357-11ec-a24f-192ff4cebbef"}', seed=0, budget=6.25), RunValue(cost=1.0, time=59.04720687866211, status=<StatusType.TIMEOUT: 2>, starttime=1650754177.6820188, endtime=1650754237.8177464, additional_info={'error': 'Timeout', 'configuration_origin': 'Initial design'})), 
(RunKey(config_id=8, instance_id='{"task_id": "9a9e1fbd-c357-11ec-a24f-192ff4cebbef"}', seed=0, budget=6.25), RunValue(cost=0.037499999999999985, time=59.08307766914368, status=<StatusType.DONOTADVANCE: 7>, starttime=1650754177.7040098, endtime=1650754237.8338203, additional_info={'duration': 57.0889732837677, 'num_run': 9, 'train_loss': 0.0353125, 'subprocess_stdout': 'The least populated class in y has only 2 members, which is less than n_splits=5.\nThe least populated class in y has only 2 members, which is less than n_splits=5.\nThe least populated class in y has only 2 members, which is less than n_splits=5.\nThe least populated class in y has only 2 members, which is less than n_splits=5.\nThe least populated class in y has only 2 members, which is less than n_splits=5.\n', 'info': 'Run stopped because of timeout.', 'learning_curve': [0.0425, 0.05124999999999999, 0.037499999999999985], 'learning_curve_runtime': [34.96041703224182, 45.19434571266174, 57.0889732837677], 'train_learning_curve': [0.03656249999999999, 0.04406250000000001, 0.0353125], 'configuration_origin': 'Initial design'})), 
(RunKey(config_id=9, instance_id='{"task_id": "9a9e1fbd-c357-11ec-a24f-192ff4cebbef"}', seed=0, budget=6.25), RunValue(cost=0.0024999999999999914, time=46.72268342971802, status=<StatusType.SUCCESS: 1>, starttime=1650754177.74943, endtime=1650754224.496036, additional_info={'duration': 46.21944880485535, 'num_run': 10, 'train_loss': 0.0, 'learning_curve': [0.003749999999999987, 0.0024999999999999914, 0.0024999999999999914, 0.0024999999999999914, 0.0024999999999999914], 'learning_curve_runtime': [22.935990571975708, 29.194799423217773, 34.853703022003174, 40.19394636154175, 46.21944880485535], 'train_learning_curve': [0.0006250000000000089, 0.0, 0.0, 0.0, 0.0], 'configuration_origin': 'Initial design'})), 
(RunKey(config_id=10, instance_id='{"task_id": "9a9e1fbd-c357-11ec-a24f-192ff4cebbef"}', seed=0, budget=6.25), RunValue(cost=1.0, time=16.99842596054077, status=<StatusType.CRASHED: 3>, starttime=1650754177.88176, endtime=1650754194.9761496, additional_info={'traceback': 'Traceback (most recent call last):\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/evaluation/__init__.py", line 42, in fit_predict_try_except_decorator\n    return ta(queue=queue, **kwargs)\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/evaluation/train_evaluator.py", line 1408, in eval_iterative_cv\n    eval_cv(\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/evaluation/train_evaluator.py", line 1383, in eval_cv\n    evaluator.fit_predict_and_loss(iterative=iterative)\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/evaluation/train_evaluator.py", line 329, in fit_predict_and_loss\n    model.iterative_fit(\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/pipeline/base.py", line 127, in iterative_fit\n    self._final_estimator.iterative_fit(X, y, n_iter=n_iter,\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/pipeline/components/classification/__init__.py", line 149, in iterative_fit\n    return self.choice.iterative_fit(X, y, n_iter=n_iter, **fit_params)\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/pipeline/components/classification/gradient_boosting.py", line 125, in iterative_fit\n    self.estimator.fit(X, y, sample_weight=sample_weight)\n  File "/home/my/.local/lib/python3.8/site-packages/sklearn/ensemble/_hist_gradient_boosting/gradient_boosting.py", line 270, in fit\n    sample_weight_val) = train_test_split(\n  File "/home/my/.local/lib/python3.8/site-packages/sklearn/model_selection/_split.py", line 2197, in train_test_split\n    train, test = next(cv.split(X=arrays[0], y=stratify))\n  File "/home/my/.local/lib/python3.8/site-packages/sklearn/model_selection/_split.py", line 1387, in split\n    for train, test in self._iter_indices(X, y, groups):\n  File "/home/my/.local/lib/python3.8/site-packages/sklearn/model_selection/_split.py", line 1715, in _iter_indices\n    raise ValueError("The least populated class in y has only 1"\nValueError: The least populated class in y has only 1 member, which is too few. The minimum number of groups for any class cannot be less than 2.\n', 'error': "ValueError('The least populated class in y has only 1 member, which is too few. The minimum number of groups for any class cannot be less than 2.')", 'configuration_origin': 'Initial design'})), 
(RunKey(config_id=11, instance_id='{"task_id": "9a9e1fbd-c357-11ec-a24f-192ff4cebbef"}', seed=0, budget=6.25), RunValue(cost=0.0024999999999999914, time=51.11537718772888, status=<StatusType.SUCCESS: 1>, starttime=1650754177.971318, endtime=1650754229.162319, additional_info={'duration': 50.44809103012085, 'num_run': 12, 'train_loss': 0.002500000000000036, 'learning_curve': [0.0024999999999999914, 0.0024999999999999914, 0.0024999999999999914, 0.0024999999999999914, 0.0024999999999999914, 0.0024999999999999914], 'learning_curve_runtime': [16.837047815322876, 22.513781547546387, 29.39050579071045, 35.171164751052856, 41.45954966545105, 50.44809103012085], 'train_learning_curve': [0.002500000000000036, 0.002500000000000036, 0.002500000000000036, 0.002500000000000036, 0.002500000000000036, 0.002500000000000036], 'configuration_origin': 'Initial design'})), 
(RunKey(config_id=12, instance_id='{"task_id": "9a9e1fbd-c357-11ec-a24f-192ff4cebbef"}', seed=0, budget=6.25), RunValue(cost=1.0, time=29.34589171409607, status=<StatusType.CRASHED: 3>, starttime=1650754178.4252083, endtime=1650754207.8164613, additional_info={'traceback': 'Traceback (most recent call last):\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/evaluation/__init__.py", line 42, in fit_predict_try_except_decorator\n    return ta(queue=queue, **kwargs)\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/evaluation/train_evaluator.py", line 1408, in eval_iterative_cv\n    eval_cv(\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/evaluation/train_evaluator.py", line 1383, in eval_cv\n    evaluator.fit_predict_and_loss(iterative=iterative)\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/evaluation/train_evaluator.py", line 329, in fit_predict_and_loss\n    model.iterative_fit(\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/pipeline/base.py", line 127, in iterative_fit\n    self._final_estimator.iterative_fit(X, y, n_iter=n_iter,\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/pipeline/components/classification/__init__.py", line 149, in iterative_fit\n    return self.choice.iterative_fit(X, y, n_iter=n_iter, **fit_params)\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/pipeline/components/classification/gradient_boosting.py", line 125, in iterative_fit\n    self.estimator.fit(X, y, sample_weight=sample_weight)\n  File "/home/my/.local/lib/python3.8/site-packages/sklearn/ensemble/_hist_gradient_boosting/gradient_boosting.py", line 270, in fit\n    sample_weight_val) = train_test_split(\n  File "/home/my/.local/lib/python3.8/site-packages/sklearn/model_selection/_split.py", line 2197, in train_test_split\n    train, test = next(cv.split(X=arrays[0], y=stratify))\n  File "/home/my/.local/lib/python3.8/site-packages/sklearn/model_selection/_split.py", line 1387, in split\n    for train, test in self._iter_indices(X, y, groups):\n  File "/home/my/.local/lib/python3.8/site-packages/sklearn/model_selection/_split.py", line 1715, in _iter_indices\n    raise ValueError("The least populated class in y has only 1"\nValueError: The least populated class in y has only 1 member, which is too few. The minimum number of groups for any class cannot be less than 2.\n', 'error': "ValueError('The least populated class in y has only 1 member, which is too few. The minimum number of groups for any class cannot be less than 2.')", 'configuration_origin': 'Initial design'})), 
(RunKey(config_id=13, instance_id='{"task_id": "9a9e1fbd-c357-11ec-a24f-192ff4cebbef"}', seed=0, budget=6.25), RunValue(cost=1.0, time=16.431341409683228, status=<StatusType.CRASHED: 3>, starttime=1650754195.1183324, endtime=1650754211.6001222, additional_info={'traceback': 'Traceback (most recent call last):\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/evaluation/__init__.py", line 42, in fit_predict_try_except_decorator\n    return ta(queue=queue, **kwargs)\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/evaluation/train_evaluator.py", line 1408, in eval_iterative_cv\n    eval_cv(\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/evaluation/train_evaluator.py", line 1383, in eval_cv\n    evaluator.fit_predict_and_loss(iterative=iterative)\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/evaluation/train_evaluator.py", line 329, in fit_predict_and_loss\n    model.iterative_fit(\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/pipeline/base.py", line 127, in iterative_fit\n    self._final_estimator.iterative_fit(X, y, n_iter=n_iter,\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/pipeline/components/classification/__init__.py", line 149, in iterative_fit\n    return self.choice.iterative_fit(X, y, n_iter=n_iter, **fit_params)\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/pipeline/components/classification/gradient_boosting.py", line 125, in iterative_fit\n    self.estimator.fit(X, y, sample_weight=sample_weight)\n  File "/home/my/.local/lib/python3.8/site-packages/sklearn/ensemble/_hist_gradient_boosting/gradient_boosting.py", line 270, in fit\n    sample_weight_val) = train_test_split(\n  File "/home/my/.local/lib/python3.8/site-packages/sklearn/model_selection/_split.py", line 2197, in train_test_split\n    train, test = next(cv.split(X=arrays[0], y=stratify))\n  File "/home/my/.local/lib/python3.8/site-packages/sklearn/model_selection/_split.py", line 1387, in split\n    for train, test in self._iter_indices(X, y, groups):\n  File "/home/my/.local/lib/python3.8/site-packages/sklearn/model_selection/_split.py", line 1715, in _iter_indices\n    raise ValueError("The least populated class in y has only 1"\nValueError: The least populated class in y has only 1 member, which is too few. The minimum number of groups for any class cannot be less than 2.\n', 'error': "ValueError('The least populated class in y has only 1 member, which is too few. The minimum number of groups for any class cannot be less than 2.')", 'configuration_origin': 'Initial design'})), 
(RunKey(config_id=14, instance_id='{"task_id": "9a9e1fbd-c357-11ec-a24f-192ff4cebbef"}', seed=0, budget=6.25), RunValue(cost=1.0, time=59.04702353477478, status=<StatusType.TIMEOUT: 2>, starttime=1650754201.3973906, endtime=1650754261.4712338, additional_info={'error': 'Timeout', 'configuration_origin': 'Initial design'})), 
(RunKey(config_id=15, instance_id='{"task_id": "9a9e1fbd-c357-11ec-a24f-192ff4cebbef"}', seed=0, budget=6.25), RunValue(cost=1.0, time=59.046088457107544, status=<StatusType.TIMEOUT: 2>, starttime=1650754207.8768454, endtime=1650754267.9492505, additional_info={'error': 'Timeout', 'configuration_origin': 'Initial design'})), 
(RunKey(config_id=16, instance_id='{"task_id": "9a9e1fbd-c357-11ec-a24f-192ff4cebbef"}', seed=0, budget=6.25), RunValue(cost=0.0024999999999999914, time=56.79266715049744, status=<StatusType.SUCCESS: 1>, starttime=1650754210.2605114, endtime=1650754267.0842903, additional_info={'duration': 56.431211709976196, 'num_run': 17, 'train_loss': 0.0, 'learning_curve': [0.003749999999999987, 0.003749999999999987, 0.003749999999999987, 0.0024999999999999914, 0.0024999999999999914], 'learning_curve_runtime': [34.58225774765015, 39.83625030517578, 45.36500835418701, 50.46578550338745, 56.431211709976196], 'train_learning_curve': [0.0009374999999999912, 0.0, 0.0, 0.0, 0.0], 'configuration_origin': 'Initial design'})), 
(RunKey(config_id=17, instance_id='{"task_id": "9a9e1fbd-c357-11ec-a24f-192ff4cebbef"}', seed=0, budget=6.25), RunValue(cost=1.0, time=59.04437851905823, status=<StatusType.TIMEOUT: 2>, starttime=1650754211.654649, endtime=1650754271.7208927, additional_info={'error': 'Timeout', 'configuration_origin': 'Initial design'})), 
(RunKey(config_id=18, instance_id='{"task_id": "9a9e1fbd-c357-11ec-a24f-192ff4cebbef"}', seed=0, budget=6.25), RunValue(cost=1.0, time=59.05082106590271, status=<StatusType.TIMEOUT: 2>, starttime=1650754224.535896, endtime=1650754284.617626, additional_info={'error': 'Timeout', 'configuration_origin': 'Initial design'})), 
(RunKey(config_id=19, instance_id='{"task_id": "9a9e1fbd-c357-11ec-a24f-192ff4cebbef"}', seed=0, budget=6.25), RunValue(cost=0.9975, time=55.06376552581787, status=<StatusType.DONOTADVANCE: 7>, starttime=1650754229.2518373, endtime=1650754285.3876128, additional_info={'duration': 48.31077289581299, 'num_run': 20, 'train_loss': 0.9975, 'subprocess_stdout': 'The least populated class in y has only 2 members, which is less than n_splits=5.\nThe least populated class in y has only 2 members, which is less than n_splits=5.\nThe least populated class in y has only 2 members, which is less than n_splits=5.\nThe least populated class in y has only 2 members, which is less than n_splits=5.\nThe least populated class in y has only 2 members, which is less than n_splits=5.\nThe least populated class in y has only 2 members, which is less than n_splits=5.\n', 'subprocess_stderr': '/home/my/.local/lib/python3.8/site-packages/sklearn/linear_model/_stochastic_gradient.py:574: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n  warnings.warn("Maximum number of iteration reached before "\n/home/my/.local/lib/python3.8/site-packages/sklearn/linear_model/_stochastic_gradient.py:574: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n  warnings.warn("Maximum number of iteration reached before "\n/home/my/.local/lib/python3.8/site-packages/sklearn/linear_model/_stochastic_gradient.py:574: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n  warnings.warn("Maximum number of iteration reached before "\n/home/my/.local/lib/python3.8/site-packages/sklearn/linear_model/_stochastic_gradient.py:574: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n  warnings.warn("Maximum number of iteration reached before "\n/home/my/.local/lib/python3.8/site-packages/sklearn/linear_model/_stochastic_gradient.py:574: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n  warnings.warn("Maximum number of iteration reached before "\n', 'info': 'Run stopped because of timeout.', 'learning_curve': [0.9975, 0.9975, 0.9975, 0.9975], 'learning_curve_runtime': [23.58888077735901, 31.481527090072632, 39.974218130111694, 48.31077289581299], 'train_learning_curve': [0.9975, 0.9975, 0.9975, 0.9975], 'configuration_origin': 'Initial design'})), 
(RunKey(config_id=20, instance_id='{"task_id": "9a9e1fbd-c357-11ec-a24f-192ff4cebbef"}', seed=0, budget=6.25), RunValue(cost=1.0, time=47.03621745109558, status=<StatusType.TIMEOUT: 2>, starttime=1650754237.4181778, endtime=1650754285.515705, additional_info={'error': 'Timeout', 'configuration_origin': 'Initial design'})), 
(RunKey(config_id=21, instance_id='{"task_id": "9a9e1fbd-c357-11ec-a24f-192ff4cebbef"}', seed=0, budget=6.25), RunValue(cost=0.12, time=47.034883975982666, status=<StatusType.DONOTADVANCE: 7>, starttime=1650754237.569879, endtime=1650754285.6265893, additional_info={'duration': 46.12961411476135, 'num_run': 22, 'train_loss': 0.002812499999999996, 'subprocess_stdout': 'The least populated class in y has only 2 members, which is less than n_splits=5.\nThe least populated class in y has only 2 members, which is less than n_splits=5.\nThe least populated class in y has only 2 members, which is less than n_splits=5.\nThe least populated class in y has only 2 members, which is less than n_splits=5.\nThe least populated class in y has only 2 members, which is less than n_splits=5.\nThe least populated class in y has only 2 members, which is less than n_splits=5.\nThe least populated class in y has only 2 members, which is less than n_splits=5.\n', 'subprocess_stderr': '/home/my/.local/lib/python3.8/site-packages/sklearn/linear_model/_stochastic_gradient.py:574: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n  warnings.warn("Maximum number of iteration reached before "\n/home/my/.local/lib/python3.8/site-packages/sklearn/linear_model/_stochastic_gradient.py:574: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n  warnings.warn("Maximum number of iteration reached before "\n/home/my/.local/lib/python3.8/site-packages/sklearn/linear_model/_stochastic_gradient.py:574: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n  warnings.warn("Maximum number of iteration reached before "\n/home/my/.local/lib/python3.8/site-packages/sklearn/linear_model/_stochastic_gradient.py:574: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n  warnings.warn("Maximum number of iteration reached before "\n/home/my/.local/lib/python3.8/site-packages/sklearn/linear_model/_stochastic_gradient.py:574: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n  warnings.warn("Maximum number of iteration reached before "\n', 'info': 'Run stopped because of timeout.', 'learning_curve': [0.25125, 0.20875000000000005, 0.19375, 0.15625000000000003, 0.12], 'learning_curve_runtime': [22.080286502838135, 27.89290499687195, 33.33868098258972, 38.730738162994385, 46.12961411476135], 'train_learning_curve': [0.15468750000000003, 0.07156250000000001, 0.02312500000000002, 0.00812499999999996, 0.002812499999999996], 'configuration_origin': 'Initial design'})), 
(RunKey(config_id=22, instance_id='{"task_id": "9a9e1fbd-c357-11ec-a24f-192ff4cebbef"}', seed=0, budget=6.25), RunValue(cost=1.0, time=20.62538194656372, status=<StatusType.CRASHED: 3>, starttime=1650754237.697609, endtime=1650754258.3523226, additional_info={'traceback': 'Traceback (most recent call last):\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/evaluation/__init__.py", line 42, in fit_predict_try_except_decorator\n    return ta(queue=queue, **kwargs)\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/evaluation/train_evaluator.py", line 1408, in eval_iterative_cv\n    eval_cv(\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/evaluation/train_evaluator.py", line 1383, in eval_cv\n    evaluator.fit_predict_and_loss(iterative=iterative)\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/evaluation/train_evaluator.py", line 329, in fit_predict_and_loss\n    model.iterative_fit(\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/pipeline/base.py", line 127, in iterative_fit\n    self._final_estimator.iterative_fit(X, y, n_iter=n_iter,\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/pipeline/components/classification/__init__.py", line 149, in iterative_fit\n    return self.choice.iterative_fit(X, y, n_iter=n_iter, **fit_params)\n  File "/home/my/anaconda3/lib/python3.8/site-packages/autosklearn/pipeline/components/classification/gradient_boosting.py", line 125, in iterative_fit\n    self.estimator.fit(X, y, sample_weight=sample_weight)\n  File "/home/my/.local/lib/python3.8/site-packages/sklearn/ensemble/_hist_gradient_boosting/gradient_boosting.py", line 270, in fit\n    sample_weight_val) = train_test_split(\n  File "/home/my/.local/lib/python3.8/site-packages/sklearn/model_selection/_split.py", line 2197, in train_test_split\n    train, test = next(cv.split(X=arrays[0], y=stratify))\n  File "/home/my/.local/lib/python3.8/site-packages/sklearn/model_selection/_split.py", line 1387, in split\n    for train, test in self._iter_indices(X, y, groups):\n  File "/home/my/.local/lib/python3.8/site-packages/sklearn/model_selection/_split.py", line 1715, in _iter_indices\n    raise ValueError("The least populated class in y has only 1"\nValueError: The least populated class in y has only 1 member, which is too few. The minimum number of groups for any class cannot be less than 2.\n', 'error': "ValueError('The least populated class in y has only 1 member, which is too few. The minimum number of groups for any class cannot be less than 2.')", 'configuration_origin': 'Initial design'})), 
(RunKey(config_id=23, instance_id='{"task_id": "9a9e1fbd-c357-11ec-a24f-192ff4cebbef"}', seed=0, budget=6.25), RunValue(cost=1.0, time=47.07687187194824, status=<StatusType.TIMEOUT: 2>, starttime=1650754238.030686, endtime=1650754286.1327865, additional_info={'error': 'Timeout', 'configuration_origin': 'Initial design'})), 
(RunKey(config_id=24, instance_id='{"task_id": "9a9e1fbd-c357-11ec-a24f-192ff4cebbef"}', seed=0, budget=6.25), RunValue(cost=1.0, time=46.08737850189209, status=<StatusType.TIMEOUT: 2>, starttime=1650754238.504264, endtime=1650754285.610624, additional_info={'error': 'Timeout', 'configuration_origin': 'Initial design'})), 
(RunKey(config_id=25, instance_id='{"task_id": "9a9e1fbd-c357-11ec-a24f-192ff4cebbef"}', seed=0, budget=6.25), RunValue(cost=0.115, time=46.0707311630249, status=<StatusType.DONOTADVANCE: 7>, starttime=1650754238.620648, endtime=1650754285.7274203, additional_info={'duration': 43.283092737197876, 'num_run': 26, 'train_loss': 0.1109375, 'subprocess_stdout': 'The least populated class in y has only 2 members, which is less than n_splits=5.\nThe least populated class in y has only 2 members, which is less than n_splits=5.\nThe least populated class in y has only 2 members, which is less than n_splits=5.\nThe least populated class in y has only 2 members, which is less than n_splits=5.\nThe least populated class in y has only 2 members, which is less than n_splits=5.\nThe least populated class in y has only 2 members, which is less than n_splits=5.\n', 'subprocess_stderr': '/home/my/.local/lib/python3.8/site-packages/sklearn/linear_model/_stochastic_gradient.py:574: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n  warnings.warn("Maximum number of iteration reached before "\n/home/my/.local/lib/python3.8/site-packages/sklearn/linear_model/_stochastic_gradient.py:574: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n  warnings.warn("Maximum number of iteration reached before "\n/home/my/.local/lib/python3.8/site-packages/sklearn/linear_model/_stochastic_gradient.py:574: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n  warnings.warn("Maximum number of iteration reached before "\n/home/my/.local/lib/python3.8/site-packages/sklearn/linear_model/_stochastic_gradient.py:574: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n  warnings.warn("Maximum number of iteration reached before "\n/home/my/.local/lib/python3.8/site-packages/sklearn/linear_model/_stochastic_gradient.py:574: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n  warnings.warn("Maximum number of iteration reached before "\n', 'info': 'Run stopped because of timeout.', 'learning_curve': [0.31375000000000003, 0.315, 0.22, 0.115], 'learning_curve_runtime': [21.78280019760132, 27.698118925094604, 34.59027099609375, 43.283092737197876], 'train_learning_curve': [0.390625, 0.37312500000000004, 0.27812500000000007, 0.1109375], 'configuration_origin': 'Initial design'})), 
(RunKey(config_id=26, instance_id='{"task_id": "9a9e1fbd-c357-11ec-a24f-192ff4cebbef"}', seed=0, budget=6.25), RunValue(cost=1.0, time=26.051013231277466, status=<StatusType.TIMEOUT: 2>, starttime=1650754258.3928027, endtime=1650754285.4660196, additional_info={'error': 'Timeout', 'configuration_origin': 'Initial design'})), 
(RunKey(config_id=27, instance_id='{"task_id": "9a9e1fbd-c357-11ec-a24f-192ff4cebbef"}', seed=0, budget=6.25), RunValue(cost=1.0, time=23.03747320175171, status=<StatusType.TIMEOUT: 2>, starttime=1650754261.5237162, endtime=1650754285.5842803, additional_info={'error': 'Timeout', 'configuration_origin': 'Initial design'})), 
(RunKey(config_id=28, instance_id='{"task_id": "9a9e1fbd-c357-11ec-a24f-192ff4cebbef"}', seed=0, budget=6.25), RunValue(cost=1.0, time=18.074486017227173, status=<StatusType.TIMEOUT: 2>, starttime=1650754267.144992, endtime=1650754286.260646, additional_info={'error': 'Timeout', 'configuration_origin': 'Initial design'})), 
(RunKey(config_id=29, instance_id='{"task_id": "9a9e1fbd-c357-11ec-a24f-192ff4cebbef"}', seed=0, budget=6.25), RunValue(cost=1.0, time=17.031450510025024, status=<StatusType.TIMEOUT: 2>, starttime=1650754267.999331, endtime=1650754286.0448165, additional_info={'error': 'Timeout', 'configuration_origin': 'Initial design'})), 
(RunKey(config_id=9, instance_id='{"task_id": "9a9e1fbd-c357-11ec-a24f-192ff4cebbef"}', seed=0, budget=25.0), RunValue(cost=1.0, time=13.043877124786377, status=<StatusType.TIMEOUT: 2>, starttime=1650754271.7868109, endtime=1650754285.8542476, additional_info={'error': 'Timeout', 'configuration_origin': 'Initial design'})), 
(RunKey(config_id=30, instance_id='{"task_id": "9a9e1fbd-c357-11ec-a24f-192ff4cebbef"}', seed=0, budget=6.25), RunValue(cost=1.0, time=0.0, status=<StatusType.STOP: 8>, starttime=1650754284.6560502, endtime=1650754284.6560507, additional_info={})), 
(RunKey(config_id=31, instance_id='{"task_id": "9a9e1fbd-c357-11ec-a24f-192ff4cebbef"}', seed=0, budget=6.25), RunValue(cost=1.0, time=0.0, status=<StatusType.STOP: 8>, starttime=1650754284.7680821, endtime=1650754284.7680821, additional_info={}))])
>>>

[eddiebergman]

So this issue is a known #1190 one and was (hopefully) fixed in development a while ago.

raise ValueError("The least populated class in y has only 1"

If you are able to, could you try the development branch?

Essentially when you have only one instance of a class, the default sklearn splitter doesn't know how to handle it when trying to split into train and test.

[asmgx]

I tried the development code that was provided in this link https://github.com/automl/auto-sklearn/pull/1250/files#diff-e0f50e611431c0f3f5a735ebea3a582b6271ea02bf758de9a82a6c46dfa3a36cR60

but still getting the same error