Cannot set up AutoML to generate reproducible results

Options
Bowser
Bowser Member Posts: 2

Here is a link to my dataset:
https://drive.google.com/file/d/12TFk-lG3YKxyZSfv5-eTQmsPDWMky-Pp/view?usp=drive_link

Here is my code:

import h2o
import pandas
h2o.init()

my_df = pd.read_csv('h2o_test.csv')
response = 'col5'
predictors= ['col1','col2','col3', 'col4']

df_test = h2o.H2OFrame(my_df.tail(1))
df_train = h2o.H2OFrame(my_df.head(my_df.shape[0]-1))
aml = h2o.automl.H2OAutoML(max_models =10, balance_classes=False,seed =1)
aml.train(predictors, response, training_frame = df_train)
aml.leader.predict(df_test)

Every time I run it, it generates different results.

PC: HP Zbook
OS: Windows 11
H2O Build project version 3.46.0.1

Any thoughts?