I tried a lot, but i can’t get better than
AssertionError: Wrong values for d[‘w’]. [[ 0.08639757]
[-0.08235296]
[-0.11799955]
[ 0.12864078]] != [[ 0.08639757]
[-0.08231268]
[-0.11798927]
[ 0.12866053]]
If i am trying to get better result on the second coordinate, i got mismatch by the 1st
Wrong values for d[‘w’]. [[ 0.08639756]
[-0.08235294]
[-0.11799953]
[ 0.12864077]] != [[ 0.08639757]
[-0.08231268]
[-0.11798927]
[ 0.12866053]]
Also i got weird result
train accuracy: 67.94258373205741
test accuracy: 84.0
Is it possible to pass this assignment actually?
Yes, it is possible to pass this assignment. Note that an error in the 4th or 5th decimal place is not a rounding error: it is a real error, indicating that something is wrong with your code. Are you sure you are handing the learning rate correctly and not hard-coding the value?
You don’t say which function is the one failing the test here. Is it optimize or model? Are you sure that your propagate code passes the tests for that function?
It would help to see the complete error trace you are getting and not just the single failing assertion.
1 Like
Holy thanks to you, i hard-coded the value, now everything is working, really thanks!