UnicodeDecodeError #16
Description
Hi
I'm using Python3.6.5 on Ubuntu 18.04. I'm able to train a classifier using:
trainer = Trainer(n_trees = 18)
trainer.fit(X_train_list, y_train)
trainer.save('fastxml.trained')
X_train_list above is a list of csr_matrix objects.
The settings file I get looks like this:
{"n_trees": 18, "max_leaf_size": 10, "max_labels_per_leaf": 20, "re_split": 0, "n_jobs": 1, "alpha": 0.0001, "seed": 2016, "n_epochs": 2, "n_updates": 100.0, "verbose": false, "bias": true, "subsample": 1, "loss": "log", "sparse_multiple": 25, "leaf_classifiers": false, "gamma": 30, "blend": 0.8, "leaf_eps": 1e-05, "optimization": "fastxml", "engine": "auto", "auto_weight": 32, "eps": 1e-06, "C": 1, "leaf_probs": false, "n_labels": 58}
However, when trying to use the inferencer I get UnicodeDecodeErrors such as this:
clf = Inferencer('fastxml.trained')
----------------------------------------------------
Traceback (most recent call last):
File "fastxml/inferencer.pyx", line 123, in fastxml.inferencer.load_sparse
values = read_row(f, 'If')
File "fastxml/inferencer.pyx", line 107, in fastxml.inferencer.read_row
d = f.read(struct.calcsize('I'))
File "/usr/lib/python3.6/codecs.py", line 321, in decode
(result, consumed) = self._buffer_decode(data, self.errors, final)
UnicodeDecodeError: 'utf-8' codec can't decode byte 0xff in position 0: invalid start byte
Thank you.