I am very interested in your work, but my computational resources are limited, so I would like to try a smaller dataset, such as IWSLT14, with settings N=6 and M=3 or N=12, 18 and M = 6. Have you tried this configuration before? If so, how were the results?
Hi there,
I am very interested in your work, but my computational resources are limited, so I would like to try a smaller dataset, such as IWSLT14, with settings N=6 and M=3 or N=12, 18 and M = 6. Have you tried this configuration before? If so, how were the results?