Vanishing Gradient
vanishing-gradient
Gradient values become too small during backpropagation → network stops learning
Common in deep rnn → weights shrink exponentially
Solutions:
*References
*References
#ml-notes
Gradient values become too small during backpropagation → network stops learning
Common in deep rnn → weights shrink exponentially
Solutions:
#ml-notes