why the model.out content is messy code about xlearn HOT 6 OPEN

I have another problem, when I run the code:

from future import absolute_import
import xlearn as xl

ffm_model = xl.create_ffm()

ffm_model.setTrain('./data/train_ffm.txt')
ffm_model.setValidate('./data/valid_ffm.txt')

param = {'task': 'binary', 'lr': 0.2, 'lambda': 0.002, 'nthread': 8, 'epoch': 10}

ffm_model.fit(param, './ffm_model.out')

ffm_model.setTest('./data/test_ffm.txt')
ffm_model.predict('./ffm_model.out', './ffm_output.txt')

it has raise the error:
Traceback (most recent call last):
File "xlearn.py", line 3, in
import xlearn as xl
File "/home/huangzhanpeng/Download/criteo/kaggle_criteo_ctr_challenge-/xlearn.py", line 5, in
ffm_model = xl.create_ffm()
AttributeError: 'module' object has no attribute 'create_ffm'

but the code can run in jupyter notebook

from xlearn.

It's a strange error I have never met before. I will check this soon.

from xlearn.

there is another strange thing that if not delete xlearn.pyc, while create another python script, it still raise the same error

from xlearn.

@Huangzhanpeng 请问后来 'module' object has no attribute 'create_ffm' 这个问题怎么解决的，我也遇到了同样的问题，谢谢。

from xlearn.

@johnwanggithub 你能把 code 放上来吗，我帮你 check 一下。

from xlearn.

@johnwanggithub 我删除了.pyc文件后就可以了

from xlearn.