def prepare(args):
idim = 10
odim = 5
model = E2E(idim, odim, args)
batchsize = 2
x = torch.randn(batchsize, 10, idim)
ilens = [10, 9]
n_token = odim - 1
y_src = (torch.rand(batchsize, 4) * n_token % n_token).long()
y_tgt = (torch.rand(batchsize, 4) * n_token % n_token).long()
olens = [3, 4]
for i in range(batchsize):
x[i, ilens[i]:] = -1
y_tgt[i, olens[i]:] = model.ignore_id
y_src[i, olens[i]:] = model.ignore_id
data = []
for i in range(batchsize):
data.append((
"utt%d" % i,
{
"input": [{
"shape": [ilens[i], idim]
}],
"output": [{
"shape": [olens[i]]
}],
},
))
return model, x, torch.tensor(ilens), y_tgt, y_src, data
def prepare(args):
idim = 10
odim = 5
model = E2E(idim, odim, args)
batchsize = 2
ilens = [10, 9]
olens = [3, 4]
n_token = odim - 1
x = torch.randn(batchsize, max(ilens), idim)
y_src = (torch.rand(batchsize, max(olens)) * n_token % n_token).long()
y_tgt = (torch.rand(batchsize, max(olens)) * n_token % n_token).long()
for i in range(batchsize):
x[i, ilens[i] :] = -1
y_tgt[i, olens[i] :] = model.ignore_id
y_src[i, olens[i] :] = model.ignore_id
data = {}
uttid_list = []
for i in range(batchsize):
data["utt%d" % i] = {
"input": [{"shape": [ilens[i], idim]}],
"output": [{"shape": [olens[i]]}],
}
uttid_list.append("utt%d" % i)
return model, x, torch.tensor(ilens), y_tgt, y_src, data, uttid_list
def add_arguments(parser):
"""Add arguments."""
E2ETransformer.add_arguments(parser)
E2E.add_conformer_arguments(parser)
return parser