Hello, with the pipeline object, is it possible to perform inferences with my 2 gpus at the same time ?
What I would like is something like:
out = pipe(
input,
batch_size=batch_size,
n_gpus=2 # <- Is there an equivalent to this argument ?
)