likejazz/Siamese-LSTM

When I run train.py I meet this error

Juary88 opened this issue · 2 comments

InternalError: Blas GEMM launch failed : a.shape=(1024, 64), b.shape=(64, 50), m=1024, n=50, k=64

Do you use theano or tensorflow backend?

@Juary88 tensorflow