Hi , i just stared with Pytrch , I am unable to understand why why graph is like this when why calculated weight and bias are close to assumed

[ChimeraElixir]

hi ,i have just started coding withi pytorch ,please take a look at the code , i created a dummy dataset while learning pytorch but after traing and testing the loss is showing to be like 0.007 and there are no further improvements if i inchease no of epoch , taken weight is .03 and bias is 0.9 and calculated is 0.3028 and 0.8910, the problem rises when i try to plot the predicted values against test data the

what is goiong on please explain if this is right

https://colab.research.google.com/drive/1rKr2khXLzHxMAqvoDl_jI2WAd6r_BkYm?usp=sharing

[ChimeraElixir]

there was an error in the plot_prediction code where i was giving the x as y_test rather than x_test