Hi,
Testing the code on the reported NYT params of:
Lambda1 = -2
lambda2 = 1
beta = 0.6
cannot the replicate the results shown in the paper, Rouge1 is around ~0.25 while reported is 0.41.
Is there anything im missing? How can I reproduce the reported results on NYT?
Hi,
Testing the code on the reported NYT params of:
Lambda1 = -2
lambda2 = 1
beta = 0.6
cannot the replicate the results shown in the paper, Rouge1 is around ~0.25 while reported is 0.41.
Is there anything im missing? How can I reproduce the reported results on NYT?