core dump when save_to_file

[hubutui]

I create a virtual environment with conda, and install thundersvm wheel. Here is the core dump info:

           PID: 593934 (python)
           UID: 1000 (hubutui)
           GID: 1002 (hubutui)
        Signal: 11 (SEGV)
     Timestamp: Tue 2020-06-02 19:37:04 CST (1min 7s ago)
  Command Line: python train.py
    Executable: /home/hubutui/.conda/envs/thundersvm/bin/python3.6
 Control Group: /user.slice/user-1000.slice/user@1000.service/apps.slice/apps-org.gnome.Terminal.slice/vte-spawn-6a03779d-690a-4b50-a18c-8ae861485284.scope
          Unit: user@1000.service
     User Unit: vte-spawn-6a03779d-690a-4b50-a18c-8ae861485284.scope
         Slice: user-1000.slice
     Owner UID: 1000 (hubutui)
       Boot ID: 39f5bcfa75d04afe8b858e5175043650
    Machine ID: f553124cb1ca464caf172ab914b34a3c
      Hostname: ArchLinux
       Storage: /var/lib/systemd/coredump/core.python.1000.39f5bcfa75d04afe8b858e5175043650.593934.1591097824000000000000.lz4
       Message: Process 593934 (python) of user 1000 dumped core.

                Stack trace of thread 593934:
                #0  0x00007f431185e083 n/a (/home/hubutui/.conda/envs/thundersvm/lib/python3.6/site-packages/thundersvm/libthundersvm.so + 0x51083)
                #1  0x00007ffd19db4d70 n/a (n/a + 0x0)

A minimal code example to reproduce:

from thundersvm import SVC
from sklearn.datasets import load_iris
cls = SVC()
data, target = load_iris(return_X_y=True)
cls.fit(data, target)
cls.save_to_file('model')

[hubutui]

sorry, I run cross_validate, but it did not return trained estimator by default.