Hi Meder, Excelent work.
I working with the package several times, but today I notice some "error"; on the test pipeline specially in the torch.testing.assert_allclose command (this function before 1.8 ignore shape when they compare- more information pytorch/pytorch#47449). Well so the correct_output always differ to the output in the test, but maybe that was take in consideration?. In fact, is this a error? or that issue was taken into account in the test pipeline?.
Greetings,
Jeffri
Hi Meder, Excelent work.
I working with the package several times, but today I notice some "error"; on the test pipeline specially in the torch.testing.assert_allclose command (this function before 1.8 ignore shape when they compare- more information pytorch/pytorch#47449). Well so the correct_output always differ to the output in the test, but maybe that was take in consideration?. In fact, is this a error? or that issue was taken into account in the test pipeline?.
Greetings,
Jeffri