Led "w" to a "nan" when change "stop" param of "trX" in 01_linear_regression

[huang-xu]

In 01_linear_regression.ipyn, when change the "stop" parameter of np.linspace in line trX = np.linspace(-1, 1, 101) to a larger number like 20 (i.e. trX = np.linspace(-1, 20, 101)), then the learned weight w will be nan.

[akashgoel1197]

sir you have to change the learning rate to 0.00001 for trX = np.linspace(-1, 20, 101) , then the learned weight will not be nan.