Sigmoid neuron for non linearly separable data

I was running the notebook 0318_FeedForwardNetwork_new-1553058835338 which is the SN application on non linealy separable data. I understand that SN fails for non linearly separable data. My doubt is- how it is possible that I am getting results different from the lecture. (Pl see image below) while I am running the same notebook downloaded from the lecture. please clarify this.

Hi, Can you share a link to your notebook?

Just a check: Can you make sure that np.random seed is same as used in the course lectures?