Look like this model not support Tensor Parallel or Pipeline Parallel, pretty bad, does we have some way to make it?
· Sign up or log in to comment
