[BUG]Segmentation fault when using FastAI neural net in AutoGluon

**Bug Report Checklist**
- [x] I provided code that demonstrates a minimal reproducible example.
- [ ] I confirmed bug exists on the latest mainline of AutoGluon via source install.
- [x] I confirmed bug exists on the latest stable version of AutoGluon.

**Describe the bug**
When using AutoGluon with the FastAI neural network, I encounter a segmentation fault. The issue seems related to FastAI’s integration since other models (e.g., XGBoost, LightGBM) work fine.

**Expected behavior**
AutoGluon should run without crashing when using the FastAI neural network.

**To Reproduce**
~~~py
predictor: TabularPredictor = TabularPredictor(
label="target",
problem_type="binary",
sample_weight="sample_weight"
eval_metric="average_precision",
verbosity=4,
)

predictor.fit(
train_data=train_data,
tuning_data=val_data,
presets="best_quality",
time_limit=5 * 60 * 60,
use_bag_holdout=True,
refit_full=False,
set_best_to_refit_full=False,
save_bag_folds=True,
ag_args_fit={"num_gpus": 1},
keep_only_best=True,
hyperparameters={'FASTAI': {}},
)
~~~

**Screenshots / Logs**
~~~
Verbosity: 4 (Maximum Logging)
=================== System Info ===================
AutoGluon Version: 1.2
Python Version: 3.12.9
Operating System: Linux
Platform Machine: x86_64
Platform Version: #1-NixOS SMP PREEMPT_DYNAMIC Fri Nov 8 15:28:28 UTC 2024
CPU Count: 12
GPU Count: 1
Memory Avail: 9.91 GB / 15.35 GB (64.5%)
Disk Space Avail: 129.24 GB / 451.26 GB (28.6%)
===================================================
Presets specified: ['best_quality']
============ fit kwarg info ============
User Specified kwargs:
{'ag_args_fit': {'num_gpus': 1},
'auto_stack': True,
'keep_only_best': True,
'num_bag_sets': 1,
'refit_full': False,
'save_bag_folds': True,
'set_best_to_refit_full': False,
'use_bag_holdout': True}
Full kwargs:
{'_feature_generator_kwargs': None,
'_save_bag_folds': None,
'ag_args': None,
'ag_args_ensemble': None,
'ag_args_fit': {'num_gpus': 1},
'auto_stack': True,
'calibrate': 'auto',
'delay_bag_sets': False,
'ds_args': {'clean_up_fits': True,
'detection_time_frac': 0.25,
'enable_callbacks': False,
'enable_ray_logging': True,
'holdout_data': None,
'holdout_frac': 0.1111111111111111,
'memory_safe_fits': True,
'n_folds': 2,
'n_repeats': 1,
'validation_procedure': 'holdout'},
'excluded_model_types': None,
'feature_generator': 'auto',
'feature_prune_kwargs': None,
'holdout_frac': None,
'hyperparameter_tune_kwargs': None,
'included_model_types': None,
'keep_only_best': True,
'learning_curves': False,
'name_suffix': None,
'num_bag_folds': None,
'num_bag_sets': 1,
'num_stack_levels': None,
'pseudo_data': None,
'raise_on_no_models_fitted': True,
'refit_full': False,
'save_bag_folds': True,
'save_space': False,
'set_best_to_refit_full': False,
'test_data': None,
'unlabeled_data': None,
'use_bag_holdout': True,
'verbosity': 4}
========================================
Setting dynamic_stacking from 'auto' to False. Reason: Skip dynamic_stacking when use_bag_holdout is enabled. (use_bag_holdout=True)
Stack configuration (auto_stack=True): num_stack_levels=1, num_bag_folds=8, num_bag_sets=1
Saving /home/yehorkhod/Desktop/conflict-forecast-private/main/auto_ml/AutogluonModels/conflictforecast_data/automl/historical_&_news_val_3_sw_10_os_F_time_start_2025-04-09_15-26-02_year_2018/learner.pkl
Saving /home/yehorkhod/Desktop/conflict-forecast-private/main/auto_ml/AutogluonModels/conflictforecast_data/automl/historical_&_news_val_3_sw_10_os_F_time_start_2025-04-09_15-26-02_year_2018/predictor.pkl
Values in column 'sample_weight' used as sample weights instead of predictive features. Evaluation metrics will ignore sample weights, specify weight_evaluation=True to instead report weighted metrics.
Beginning AutoGluon training ... Time limit = 18000s
AutoGluon will save models to "/home/yehorkhod/Desktop/conflict-forecast-private/main/auto_ml/AutogluonModels/conflictforecast_data/automl/historical_&_news_val_3_sw_10_os_F_time_start_2025-04-09_15-26-02_year_2018"
Train Data Rows: 7435
Train Data Columns: 23
Tuning Data Rows: 5444
Tuning Data Columns: 23
Label Column: target
Problem Type: binary
Preprocessing data ...
Selected class <--> label mapping: class 1 = 1, class 0 = 0
Using Feature Generators to preprocess the data ...
Fitting AutoMLPipelineFeatureGenerator...
Available Memory: 10091.96 MB
Train Data (Original) Memory Usage: 2.16 MB (0.0% of available memory)
Inferring data type of each feature based on column values. Set feature_metadata_in to manually specify special dtypes of the features.
Stage 1 Generators:
Fitting AsTypeFeatureGenerator...
Original Features (exact raw dtype, raw dtype):
('float64', 'float') : 15 | ['discounted_anyviolence', 'discounted_armedconf', 'discounted_best', 'past_bestpc_12', 'past_bestpc_120', ...]
('int64', 'int') : 7 | ['ongoing_anyviolence', 'since_anyviolence', 'since_armedconf', 'since_civilwar', 'population', ...]
Types of features in original data (raw dtype, special dtypes):
('float', []) : 15 | ['discounted_anyviolence', 'discounted_armedconf', 'discounted_best', 'past_bestpc_12', 'past_bestpc_120', ...]
('int', []) : 7 | ['ongoing_anyviolence', 'since_anyviolence', 'since_armedconf', 'since_civilwar', 'population', ...]
Types of features in processed data (exact raw dtype, raw dtype):
('float64', 'float') : 15 | ['discounted_anyviolence', 'discounted_armedconf', 'discounted_best', 'past_bestpc_12', 'past_bestpc_120', ...]
('int64', 'int') : 7 | ['ongoing_anyviolence', 'since_anyviolence', 'since_armedconf', 'since_civilwar', 'population', ...]
Types of features in processed data (raw dtype, special dtypes):
('float', []) : 15 | ['discounted_anyviolence', 'discounted_armedconf', 'discounted_best', 'past_bestpc_12', 'past_bestpc_120', ...]
('int', []) : 7 | ['ongoing_anyviolence', 'since_anyviolence', 'since_armedconf', 'since_civilwar', 'population', ...]
0.0s = Fit runtime
22 features in original data used to generate 22 features in processed data.
Stage 2 Generators:
Fitting FillNaFeatureGenerator...
Types of features in original data (raw dtype, special dtypes):
('float', []) : 15 | ['discounted_anyviolence', 'discounted_armedconf', 'discounted_best', 'past_bestpc_12', 'past_bestpc_120', ...]
('int', []) : 7 | ['ongoing_anyviolence', 'since_anyviolence', 'since_armedconf', 'since_civilwar', 'population', ...]
Types of features in processed data (exact raw dtype, raw dtype):
('float64', 'float') : 15 | ['discounted_anyviolence', 'discounted_armedconf', 'discounted_best', 'past_bestpc_12', 'past_bestpc_120', ...]
('int64', 'int') : 7 | ['ongoing_anyviolence', 'since_anyviolence', 'since_armedconf', 'since_civilwar', 'population', ...]
Types of features in processed data (raw dtype, special dtypes):
('float', []) : 15 | ['discounted_anyviolence', 'discounted_armedconf', 'discounted_best', 'past_bestpc_12', 'past_bestpc_120', ...]
('int', []) : 7 | ['ongoing_anyviolence', 'since_anyviolence', 'since_armedconf', 'since_civilwar', 'population', ...]
0.0s = Fit runtime
22 features in original data used to generate 22 features in processed data.
Stage 3 Generators:
Fitting IdentityFeatureGenerator...
Types of features in original data (raw dtype, special dtypes):
('float', []) : 15 | ['discounted_anyviolence', 'discounted_armedconf', 'discounted_best', 'past_bestpc_12', 'past_bestpc_120', ...]
('int', []) : 7 | ['ongoing_anyviolence', 'since_anyviolence', 'since_armedconf', 'since_civilwar', 'population', ...]
Types of features in processed data (exact raw dtype, raw dtype):
('float64', 'float') : 15 | ['discounted_anyviolence', 'discounted_armedconf', 'discounted_best', 'past_bestpc_12', 'past_bestpc_120', ...]
('int64', 'int') : 7 | ['ongoing_anyviolence', 'since_anyviolence', 'since_armedconf', 'since_civilwar', 'population', ...]
Types of features in processed data (raw dtype, special dtypes):
('float', []) : 15 | ['discounted_anyviolence', 'discounted_armedconf', 'discounted_best', 'past_bestpc_12', 'past_bestpc_120', ...]
('int', []) : 7 | ['ongoing_anyviolence', 'since_anyviolence', 'since_armedconf', 'since_civilwar', 'population', ...]
0.0s = Fit runtime
22 features in original data used to generate 22 features in processed data.
Skipping CategoryFeatureGenerator: No input feature with required dtypes.
Skipping DatetimeFeatureGenerator: No input feature with required dtypes.
Skipping TextSpecialFeatureGenerator: No input feature with required dtypes.
Skipping TextNgramFeatureGenerator: No input feature with required dtypes.
Skipping IdentityFeatureGenerator: No input feature with required dtypes.
Skipping IsNanFeatureGenerator: No input feature with required dtypes.
Stage 4 Generators:
Fitting DropUniqueFeatureGenerator...
Types of features in original data (raw dtype, special dtypes):
('float', []) : 15 | ['discounted_anyviolence', 'discounted_armedconf', 'discounted_best', 'past_bestpc_12', 'past_bestpc_120', ...]
('int', []) : 7 | ['ongoing_anyviolence', 'since_anyviolence', 'since_armedconf', 'since_civilwar', 'population', ...]
Types of features in processed data (exact raw dtype, raw dtype):
('float64', 'float') : 15 | ['discounted_anyviolence', 'discounted_armedconf', 'discounted_best', 'past_bestpc_12', 'past_bestpc_120', ...]
('int64', 'int') : 7 | ['ongoing_anyviolence', 'since_anyviolence', 'since_armedconf', 'since_civilwar', 'population', ...]
Types of features in processed data (raw dtype, special dtypes):
('float', []) : 15 | ['discounted_anyviolence', 'discounted_armedconf', 'discounted_best', 'past_bestpc_12', 'past_bestpc_120', ...]
('int', []) : 7 | ['ongoing_anyviolence', 'since_anyviolence', 'since_armedconf', 'since_civilwar', 'population', ...]
0.0s = Fit runtime
22 features in original data used to generate 22 features in processed data.
Stage 5 Generators:
Fitting DropDuplicatesFeatureGenerator...
Types of features in original data (raw dtype, special dtypes):
('float', []) : 15 | ['discounted_anyviolence', 'discounted_armedconf', 'discounted_best', 'past_bestpc_12', 'past_bestpc_120', ...]
('int', []) : 7 | ['ongoing_anyviolence', 'since_anyviolence', 'since_armedconf', 'since_civilwar', 'population', ...]
Types of features in processed data (exact raw dtype, raw dtype):
('float64', 'float') : 15 | ['discounted_anyviolence', 'discounted_armedconf', 'discounted_best', 'past_bestpc_12', 'past_bestpc_120', ...]
('int64', 'int') : 7 | ['ongoing_anyviolence', 'since_anyviolence', 'since_armedconf', 'since_civilwar', 'population', ...]
Types of features in processed data (raw dtype, special dtypes):
('float', []) : 15 | ['discounted_anyviolence', 'discounted_armedconf', 'discounted_best', 'past_bestpc_12', 'past_bestpc_120', ...]
('int', []) : 7 | ['ongoing_anyviolence', 'since_anyviolence', 'since_armedconf', 'since_civilwar', 'population', ...]
0.0s = Fit runtime
22 features in original data used to generate 22 features in processed data.
Types of features in original data (exact raw dtype, raw dtype):
('float64', 'float') : 15 | ['discounted_anyviolence', 'discounted_armedconf', 'discounted_best', 'past_bestpc_12', 'past_bestpc_120', ...]
('int64', 'int') : 7 | ['ongoing_anyviolence', 'since_anyviolence', 'since_armedconf', 'since_civilwar', 'population', ...]
Types of features in original data (raw dtype, special dtypes):
('float', []) : 15 | ['discounted_anyviolence', 'discounted_armedconf', 'discounted_best', 'past_bestpc_12', 'past_bestpc_120', ...]
('int', []) : 7 | ['ongoing_anyviolence', 'since_anyviolence', 'since_armedconf', 'since_civilwar', 'population', ...]
Types of features in processed data (exact raw dtype, raw dtype):
('float64', 'float') : 15 | ['discounted_anyviolence', 'discounted_armedconf', 'discounted_best', 'past_bestpc_12', 'past_bestpc_120', ...]
('int64', 'int') : 7 | ['ongoing_anyviolence', 'since_anyviolence', 'since_armedconf', 'since_civilwar', 'population', ...]
Types of features in processed data (raw dtype, special dtypes):
('float', []) : 15 | ['discounted_anyviolence', 'discounted_armedconf', 'discounted_best', 'past_bestpc_12', 'past_bestpc_120', ...]
('int', []) : 7 | ['ongoing_anyviolence', 'since_anyviolence', 'since_armedconf', 'since_civilwar', 'population', ...]
0.0s = Fit runtime
22 features in original data used to generate 22 features in processed data.
Train Data (Processed) Memory Usage: 2.16 MB (0.0% of available memory)
Data preprocessing and feature engineering runtime = 0.03s ...
AutoGluon will gauge predictive performance using evaluation metric: 'average_precision'
This metric expects predicted probabilities rather than predicted class labels, so you'll need to use predict_proba() instead of predict()
To change this, specify the eval_metric parameter of Predictor()
Saving /home/yehorkhod/Desktop/conflict-forecast-private/main/auto_ml/AutogluonModels/conflictforecast_data/automl/historical_&_news_val_3_sw_10_os_F_time_start_2025-04-09_15-26-02_year_2018/learner.pkl
use_bag_holdout=True, will use tuning_data as holdout (will not be used for early stopping).
User-specified model hyperparameters to be fit:
{
'FASTAI': [{}],
}
Saving /home/yehorkhod/Desktop/conflict-forecast-private/main/auto_ml/AutogluonModels/conflictforecast_data/automl/historical_&_news_val_3_sw_10_os_F_time_start_2025-04-09_15-26-02_year_2018/utils/data/X.pkl
Saving /home/yehorkhod/Desktop/conflict-forecast-private/main/auto_ml/AutogluonModels/conflictforecast_data/automl/historical_&_news_val_3_sw_10_os_F_time_start_2025-04-09_15-26-02_year_2018/utils/data/y.pkl
Saving /home/yehorkhod/Desktop/conflict-forecast-private/main/auto_ml/AutogluonModels/conflictforecast_data/automl/historical_&_news_val_3_sw_10_os_F_time_start_2025-04-09_15-26-02_year_2018/utils/data/X_val.pkl
Saving /home/yehorkhod/Desktop/conflict-forecast-private/main/auto_ml/AutogluonModels/conflictforecast_data/automl/historical_&_news_val_3_sw_10_os_F_time_start_2025-04-09_15-26-02_year_2018/utils/data/y_val.pkl
AutoGluon will fit 2 stack levels (L1 to L2) ...
Model configs that will be trained (in order):
NeuralNetFastAI_BAG_L1: {'ag_args': {'model_type': , 'priority': 50}, 'ag_args_ensemble': {'save_bag_folds': True}, 'ag_args_fit': {'num_gpus': 1}}
Fitting 1 L1 models, fit_strategy="parallel" ...
WARNING: fit_strategy='parallel', but `num_gpus=1` is specified. GPU is not yet supported for `parallel` fit_strategy. To enable parallel, ensure you specify `num_gpus=0` in the fit call. Falling back to fit_strategy='sequential' ...
Fitting model: NeuralNetFastAI_BAG_L1 ... Training model for up to 11996.98s of the 17999.97s of remaining time.
Dropped 1 of 23 features.
Dropped 1 of 23 features.
Fitting NeuralNetFastAI_BAG_L1 with 'num_gpus': 1, 'num_cpus': 12
Saving /home/yehorkhod/Desktop/conflict-forecast-private/main/auto_ml/AutogluonModels/conflictforecast_data/automl/historical_&_news_val_3_sw_10_os_F_time_start_2025-04-09_15-26-02_year_2018/models/NeuralNetFastAI_BAG_L1/utils/model_template.pkl
Loading: /home/yehorkhod/Desktop/conflict-forecast-private/main/auto_ml/AutogluonModels/conflictforecast_data/automl/historical_&_news_val_3_sw_10_os_F_time_start_2025-04-09_15-26-02_year_2018/models/NeuralNetFastAI_BAG_L1/utils/model_template.pkl
Upper level total_num_cpus, num_gpus 12 | 1
Dropped 1 of 23 features.
minimum_model_resources: {'num_cpus': 1, 'num_gpus': 0.5}
user_cpu_per_job, user_gpu_per_job None | 1
user_ensemble_cpu, user_ensemble_gpu None | None
Resources info for GpuResourceCalculator: {'resources_per_job': {'cpu': 1, 'gpu': 1}, 'num_parallel_jobs': 1.0, 'batches': 8, 'cpu_per_job': 1, 'gpu_per_job': 1}
Fitting 8 child models (S1F1 - S1F8) | Fitting with ParallelLocalFoldFittingStrategy (1.0 workers, per: cpus=1, gpus=1, memory=0.15%)
Dispatching folds on node 9f24e01f0a8a6d06c34444b5915eaaecd1780fa336a417dd97d01186
/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/fastai/learner.py:455: UserWarning: load_learner` uses Python's insecure pickle module, which can execute malicious arbitrary code when loading. Only load files you trust.
If you only need to load model weights and optimizer state, use the safe `Learner.load` instead.
warn("load_learner` uses Python's insecure pickle module, which can execute malicious arbitrary code when loading. Only load files you trust.\nIf you only need to load model weights and optimizer state, use the safe `Learner.load` instead.")
*** SIGSEGV received at time=1744201605 on cpu 3 ***
PC: @ 0xf (unknown) (unknown)
@ 0x7f7d47389620 (unknown) (unknown)
@ 0xf (unknown) (unknown)
[2025-04-09 15:26:45,346 E 207085 207085] logging.cc:440: *** SIGSEGV received at time=1744201605 on cpu 3 ***
[2025-04-09 15:26:45,349 E 207085 207085] logging.cc:440: PC: @ 0xf (unknown) (unknown)
[2025-04-09 15:26:45,351 E 207085 207085] logging.cc:440: @ 0x7f7d47389620 (unknown) (unknown)
[2025-04-09 15:26:45,353 E 207085 207085] logging.cc:440: @ 0xf (unknown) (unknown)
Fatal Python error: Segmentation fault

Stack (most recent call first):
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/torch/cuda/__init__.py", line 319 in _lazy_init
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/torch/cuda/__init__.py", line 940 in current_device
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/fastai/torch_core.py", line 274 in default_device
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/fastai/callback/core.py", line 89 in before_fit
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/fastai/callback/core.py", line 62 in __call__
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/fastai/learner.py", line 184 in _call_one
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/fastcore/basics.py", line 919 in __call__
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/fastcore/basics.py", line 934 in map_ex
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/fastcore/foundation.py", line 163 in map
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/fastai/learner.py", line 180 in __call__
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/fastai/learner.py", line 275 in __enter__
File "/home/yehorkhod/.conda/envs/conflictai/lib/python3.12/contextlib.py", line 526 in enter_context
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/fastcore/basics.py", line 919 in __call__
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/fastcore/basics.py", line 934 in map_ex
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/fastcore/foundation.py", line 163 in map
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/fastcore/xtras.py", line 622 in __enter__
File "/home/yehorkhod/.conda/envs/conflictai/lib/python3.12/contextlib.py", line 526 in enter_context
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/fastcore/basics.py", line 919 in __call__
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/fastcore/basics.py", line 934 in map_ex
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/fastcore/foundation.py", line 163 in map
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/fastcore/xtras.py", line 622 in __enter__
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/fastai/learner.py", line 315 in get_preds
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/autogluon/tabular/models/fastainn/tabular_nn_fastai.py", line 493 in _predict_proba
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/autogluon/core/models/abstract/abstract_model.py", line 1061 in _predict_proba_internal
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/autogluon/core/models/abstract/abstract_model.py", line 1047 in predict_proba
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/autogluon/core/models/ensemble/bagged_ensemble_model.py", line 466 in _predict_proba_internal
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/autogluon/core/models/abstract/abstract_model.py", line 1047 in predict_proba
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/autogluon/core/trainer/abstract_trainer.py", line 2119 in _train_and_save
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/autogluon/core/trainer/abstract_trainer.py", line 2538 in _train_single_full
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/autogluon/core/trainer/abstract_trainer.py", line 4621 in _detached_train_multi_fold
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/autogluon/core/trainer/abstract_trainer.py", line 2922 in _train_multi_fold
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/autogluon/core/trainer/abstract_trainer.py", line 2765 in _train_multi_initial
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/autogluon/core/trainer/abstract_trainer.py", line 3170 in _train_multi
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/autogluon/core/trainer/abstract_trainer.py", line 835 in stack_new_level_core
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/autogluon/core/trainer/abstract_trainer.py", line 688 in stack_new_level
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/autogluon/core/trainer/abstract_trainer.py", line 493 in train_multi_levels
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/autogluon/core/trainer/abstract_trainer.py", line 3238 in _train_multi_and_ensemble
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/autogluon/tabular/trainer/auto_trainer.py", line 135 in fit
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/autogluon/tabular/learner/default_learner.py", line 131 in _fit
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/autogluon/tabular/learner/abstract_learner.py", line 159 in fit
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/autogluon/tabular/predictor/predictor.py", line 1305 in _fit
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/autogluon/tabular/predictor/predictor.py", line 1299 in fit
File "/home/yehorkhod/Desktop/conflict-forecast-private/.venv/lib/python3.12/site-packages/autogluon/core/utils/decorators.py", line 31 in _call
File "/home/yehorkhod/Desktop/conflict-forecast-private/main/auto_ml/cf_automl.py", line 276 in

Extension modules: numpy.core._multiarray_umath, numpy.core._multiarray_tests, numpy.linalg._umath_linalg, numpy.fft._pocketfft_internal, numpy.random._common, numpy.random.bit_generator, numpy.random._bounded_integers, numpy.random._mt19937, numpy.random.mtrand, numpy.random._philox, numpy.random._pcg64, numpy.random._sfc64, numpy.random._generator, google._upb._message, zstandard.backend_c, charset_normalizer.md, requests.packages.charset_normalizer.md, requests.packages.chardet.md, h5py._errors, h5py.defs, h5py._objects, h5py.h5, h5py.utils, h5py.h5t, h5py.h5s, h5py.h5ac, h5py.h5p, h5py.h5r, h5py._proxy, h5py._conv, h5py.h5z, h5py.h5a, h5py.h5d, h5py.h5ds, h5py.h5g, h5py.h5i, h5py.h5o, h5py.h5f, h5py.h5fd, h5py.h5pl, h5py.h5l, h5py._selector, scipy._lib._ccallback_c, scipy.sparse._sparsetools, _csparsetools, scipy.sparse._csparsetools, scipy.linalg._fblas, scipy.linalg._flapack, scipy.linalg.cython_lapack, scipy.linalg._cythonized_array_utils, scipy.linalg._solve_toeplitz, scipy.linalg._decomp_lu_cython, scipy.linalg._matfuncs_sqrtm_triu, scipy.linalg._matfuncs_expm, scipy.linalg._linalg_pythran, scipy.linalg.cython_blas, scipy.linalg._decomp_update, scipy.sparse.linalg._dsolve._superlu, scipy.sparse.linalg._eigen.arpack._arpack, scipy.sparse.linalg._propack._spropack, scipy.sparse.linalg._propack._dpropack, scipy.sparse.linalg._propack._cpropack, scipy.sparse.linalg._propack._zpropack, scipy.sparse.csgraph._tools, scipy.sparse.csgraph._shortest_path, scipy.sparse.csgraph._traversal, scipy.sparse.csgraph._min_spanning_tree, scipy.sparse.csgraph._flow, scipy.sparse.csgraph._matching, scipy.sparse.csgraph._reordering, PIL._imaging, psutil._psutil_linux, psutil._psutil_posix, pyarrow.lib, pandas._libs.tslibs.ccalendar, pandas._libs.tslibs.np_datetime, pandas._libs.tslibs.dtypes, pandas._libs.tslibs.base, pandas._libs.tslibs.nattype, pandas._libs.tslibs.timezones, pandas._libs.tslibs.fields, pandas._libs.tslibs.timedeltas, pandas._libs.tslibs.tzconversion, pandas._libs.tslibs.timestamps, pandas._libs.properties, pandas._libs.tslibs.offsets, pandas._libs.tslibs.strptime, pandas._libs.tslibs.parsing, pandas._libs.tslibs.conversion, pandas._libs.tslibs.period, pandas._libs.tslibs.vectorized, pandas._libs.ops_dispatch, pandas._libs.missing, pandas._libs.hashtable, pandas._libs.algos, pandas._libs.interval, pandas._libs.lib, pyarrow._compute, pandas._libs.ops, pandas._libs.hashing, pandas._libs.arrays, pandas._libs.tslib, pandas._libs.sparse, pandas._libs.internals, pandas._libs.indexing, pandas._libs.index, pandas._libs.writers, pandas._libs.join, pandas._libs.window.aggregations, pandas._libs.window.indexers, pandas._libs.reshape, pandas._libs.groupby, pandas._libs.json, pandas._libs.parsers, pandas._libs.testing, kiwisolver._cext, sklearn.__check_build._check_build, scipy.special._ufuncs_cxx, scipy.special._ufuncs, scipy.special._specfun, scipy.special._comb, scipy.special._ellip_harm_2, scipy.spatial._ckdtree, scipy._lib.messagestream, scipy.spatial._qhull, scipy.spatial._voronoi, scipy.spatial._distance_wrap, scipy.spatial._hausdorff, scipy.spatial.transform._rotation, scipy.optimize._group_columns, scipy.optimize._trlib._trlib, scipy.optimize._lbfgsb, _moduleTNC, scipy.optimize._moduleTNC, scipy.optimize._cobyla, scipy.optimize._slsqp, scipy.optimize._minpack, scipy.optimize._lsq.givens_elimination, scipy.optimize._zeros, scipy.optimize._cython_nnls, scipy._lib._uarray._uarray, scipy.linalg._decomp_interpolative, scipy.optimize._bglu_dense, scipy.optimize._lsap, scipy.optimize._direct, scipy.integrate._odepack, scipy.integrate._quadpack, scipy.integrate._vode, scipy.integrate._dop, scipy.integrate._lsoda, scipy.interpolate._fitpack, scipy.interpolate._dfitpack, scipy.interpolate._dierckx, scipy.interpolate._ppoly, scipy.interpolate._interpnd, scipy.interpolate._rbfinterp_pythran, scipy.interpolate._rgi_cython, scipy.interpolate._bspl, scipy.special.cython_special, scipy.stats._stats, scipy.stats._sobol, scipy.stats._qmc_cy, scipy.stats._biasedurn, scipy.stats._stats_pythran, scipy.stats._levy_stable.levyst, scipy.stats._ansari_swilk_statistics, scipy.stats._mvn, scipy.stats._rcont.rcont, scipy.ndimage._nd_image, scipy.ndimage._rank_filter_1d, _ni_label, scipy.ndimage._ni_label, sklearn.utils._isfinite, sklearn.utils.sparsefuncs_fast, sklearn.utils.murmurhash, sklearn.utils._openmp_helpers, sklearn.metrics.cluster._expected_mutual_info_fast, sklearn.preprocessing._csr_polynomial_expansion, sklearn.preprocessing._target_encoder_fast, sklearn.metrics._dist_metrics, sklearn.metrics._pairwise_distances_reduction._datasets_pair, sklearn.utils._cython_blas, sklearn.metrics._pairwise_distances_reduction._base, sklearn.metrics._pairwise_distances_reduction._middle_term_computer, sklearn.utils._heap, sklearn.utils._sorting, sklearn.metrics._pairwise_distances_reduction._argkmin, sklearn.metrics._pairwise_distances_reduction._argkmin_classmode, sklearn.utils._vector_sentinel, sklearn.metrics._pairwise_distances_reduction._radius_neighbors, sklearn.metrics._pairwise_distances_reduction._radius_neighbors_classmode, sklearn.metrics._pairwise_fast, sklearn.utils._random, sklearn.neighbors._partition_nodes, sklearn.neighbors._ball_tree, sklearn.neighbors._kd_tree, sklearn.utils.arrayfuncs, sklearn.utils._seq_dataset, sklearn.linear_model._cd_fast, _loss, sklearn._loss._loss, sklearn.svm._liblinear, sklearn.svm._libsvm, sklearn.svm._libsvm_sparse, sklearn.utils._weight_vector, sklearn.linear_model._sgd_fast, sklearn.linear_model._sag_fast, sklearn.decomposition._online_lda_fast, sklearn.decomposition._cdnmf_fast, sklearn.feature_extraction._hashing_fast, scipy.io.matlab._mio_utils, scipy.io.matlab._streams, scipy.io.matlab._mio5_utils, sklearn.datasets._svmlight_format_fast, sklearn.tree._utils, sklearn.neighbors._quad_tree, sklearn.tree._tree, sklearn.tree._splitter, sklearn.tree._criterion, sklearn.utils._fast_dict, sklearn.cluster._hierarchical_fast, sklearn.cluster._k_means_common, sklearn.cluster._k_means_elkan, sklearn.cluster._k_means_lloyd, sklearn.cluster._k_means_minibatch, sklearn.cluster._dbscan_inner, sklearn.cluster._hdbscan._tree, sklearn.cluster._hdbscan._linkage, sklearn.cluster._hdbscan._reachability, sklearn._isotonic, sklearn.manifold._barnes_hut_tsne, sklearn.manifold._utils, sklearn.ensemble._gradient_boosting, sklearn.ensemble._hist_gradient_boosting.common, sklearn.ensemble._hist_gradient_boosting._gradient_boosting, sklearn.ensemble._hist_gradient_boosting._binning, sklearn.ensemble._hist_gradient_boosting._bitset, sklearn.ensemble._hist_gradient_boosting.histogram, sklearn.ensemble._hist_gradient_boosting._predictor, sklearn.ensemble._hist_gradient_boosting.splitting, pyarrow._parquet, pyarrow._fs, pyarrow._azurefs, pyarrow._hdfs, pyarrow._gcsfs, pyarrow._s3fs, yaml._yaml, scipy.cluster._vq, scipy.cluster._hierarchy, scipy.cluster._optimal_leaf_ordering, numba.core.typeconv._typeconv, numba._helperlib, numba._dynfunc, numba._dispatcher, numba.core.typing.builtins.itertools, numba.cpython.builtins.math, numba.core.runtime._nrt_python, numba.np.ufunc._internal, numba.experimental.jitclass._box, shap._cext, PIL._imagingmath, PIL._imagingtk, torch._C, torch._C._dynamo.autograd_compiler, torch._C._dynamo.eval_frame, torch._C._dynamo.guards, torch._C._dynamo.utils, torch._C._fft, torch._C._linalg, torch._C._nested, torch._C._nn, torch._C._sparse, torch._C._special, msgpack._cmsgpack, setproctitle, ray._raylet, multidict._multidict, yarl._quoting_c, propcache._helpers_c, aiohttp._http_writer, aiohttp._http_parser, aiohttp._websocket.mask, aiohttp._websocket.reader_c, frozenlist._frozenlist, grpc._cython.cygrpc, pyarrow._json (total: 286)
~~~

**Installed Versions**

INSTALLED VERSIONS
------------------
date : 2025-04-09
time : 15:36:41.757312
python : 3.12.9.final.0
OS : Linux
OS-release : 6.6.60
Version : #1-NixOS SMP PREEMPT_DYNAMIC Fri Nov 8 15:28:28 UTC 2024
machine : x86_64
processor :
num_cores : 12
cpu_ram_mb : 15721.5546875
cuda version : 12.550.127.05
num_gpus : 1
gpu_ram_mb : [7923]
avail_disk_size_mb : 131971

accelerate : 0.34.2
autogluon : 1.2
autogluon.common : 1.2
autogluon.core : 1.2
autogluon.features : 1.2
autogluon.multimodal : 1.2
autogluon.tabular : 1.2
autogluon.timeseries : 1.2
boto3 : 1.37.29
catboost : 1.2.7
coreforecast : 0.0.12
defusedxml : 0.7.1
einops : 0.8.1
evaluate : 0.4.3
fastai : 2.7.19
fugue : 0.9.1
gluonts : 0.16.1
huggingface-hub : 0.30.2
hyperopt : 0.2.7
imodels : None
jinja2 : 3.1.6
joblib : 1.4.2
jsonschema : 4.21.1
lightgbm : 4.5.0
lightning : 2.5.1
matplotlib : 3.9.0
mlforecast : 0.13.4
networkx : 3.4.2
nlpaug : 1.1.11
nltk : 3.8.1
numpy : 1.26.4
nvidia-ml-py3 : 7.352.0
omegaconf : 2.2.3
onnx : None
onnxruntime : None
onnxruntime-gpu : None
openmim : 0.3.9
optimum : None
optimum-intel : None
orjson : 3.10.16
pandas : 2.2.3
pdf2image : 1.17.0
Pillow : 11.1.0
psutil : 6.1.1
pyarrow : 18.1.0
pytesseract : 0.3.10
pytorch-lightning : 2.5.1
pytorch-metric-learning: 2.3.0
ray : 2.39.0
requests : 2.32.3
scikit-image : 0.24.0
scikit-learn : 1.5.2
scikit-learn-intelex : None
scipy : 1.15.2
seqeval : 1.2.2
skl2onnx : None
spacy : 3.7.5
statsforecast : 1.7.8
tabpfn : 2.0.8
tensorboard : 2.19.0
text-unidecode : 1.3
timm : 1.0.3
torch : 2.5.1
torchmetrics : 1.2.1
torchvision : 0.20.1
tqdm : 4.67.1
transformers : 4.51.1
utilsforecast : 0.2.4
vowpalwabbit : None
xgboost : 2.1.4

Start by reproducing the provided TabularPredictor.fit configuration with hyperparameters set to FASTAI and one GPU. Inspect the FastAI integration around the failing training path; done means the configuration completes without a segmentation fault while the other model types continue to work.

Written by the indexing model from the issue text.