Open
Description
What is wrong?
recurrent.LSTM() outputs a string instead of an array with array based training data, this just seems like a major design flaw to me
Where does it happen?
running a basic instance of recurrent.lstm()
How do we replicate the issue?
- Step 1 create a basic recurrent.lstm() instance
- Step 2 trained it with incredibly basic data cus i thought i was going insane: net.train([{input:[1,2,3], output: [4,5,6]},{input:[4,5,6], output: [7,8,9]}], {
iterations:100,
log:true,
errorThresh:0.0001
}); - Step 3 i logged the typeof net.run([1,2,3]), and got a string. dafuq?????????????????????????????????????????????????????
Expected behavior (i.e. solution)
get an array that i can run methods on and easily get the values of