Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

EURO: small fix (temporarily remove support for nbest_rescoring) #4875

Merged
merged 1 commit into from
Jan 18, 2023
Merged

EURO: small fix (temporarily remove support for nbest_rescoring) #4875

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Jump to
Jump to file
Failed to load files.
Diff view
Diff view
89 changes: 3 additions & 86 deletions espnet2/bin/uasr_inference_k2.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -105,7 +105,6 @@ def __init__(
assert check_argument_types()

# 1. Build UASR model
logging.info(f"==========device to build model from: {device}===========")
uasr_model, uasr_train_args = UASRTask.build_model_from_file(
uasr_train_config, uasr_model_file, device
)
Expand Down Expand Up @@ -252,93 +251,11 @@ def __call__(
lattices.scores *= self.lattice_weight

results = []
# TODO(Dongji): add nbest_rescoring
if self.use_nbest_rescoring:
(
am_scores,
lm_scores,
token_ids,
new2old,
path_to_seq_map,
seq_to_path_splits,
) = nbest_am_lm_scores(
lattices, self.num_paths, self.device, self.nbest_batch_size
raise ValueError(
"Currently nbest rescoring is not supported"
)

ys_pad_lens = torch.tensor([len(hyp) for hyp in token_ids]).to(self.device)
max_token_length = max(ys_pad_lens)
ys_pad_list = []
for hyp in token_ids:
ys_pad_list.append(
torch.cat(
[
torch.tensor(hyp, dtype=torch.long),
torch.tensor(
[self.uasr_model_ignore_id]
* (max_token_length.item() - len(hyp)),
dtype=torch.long,
),
]
)
)

ys_pad = (
torch.stack(ys_pad_list).to(torch.long).to(self.device)
) # [batch, max_token_length]

encoder_out = generated_sample.index_select(
0, path_to_seq_map.to(torch.long)
).to(
self.device
) # [batch, T, dim]
encoder_out_lens = encoder_out_lens.index_select(
0, path_to_seq_map.to(torch.long)
).to(
self.device
) # [batch]

decoder_scores = -self.uasr_model.batchify_nll(
encoder_out, encoder_out_lens, ys_pad, ys_pad_lens, self.nll_batch_size
)

# padded_value for nnlm is 0
ys_pad[ys_pad == self.uasr_model_ignore_id] = 0
nnlm_nll, x_lengths = self.lm.batchify_nll(
ys_pad, ys_pad_lens, self.nll_batch_size
)
nnlm_scores = -nnlm_nll.sum(dim=1)

batch_tot_scores = (
self.am_weight * am_scores
+ self.decoder_weight * decoder_scores
+ self.nnlm_weight * nnlm_scores
)
split_size = indices_to_split_size(
seq_to_path_splits.tolist(), total_elements=batch_tot_scores.size(0)
)
batch_tot_scores = torch.split(
batch_tot_scores,
split_size,
)

hyps = []
scores = []
processed_seqs = 0
for tot_scores in batch_tot_scores:
if tot_scores.nelement() == 0:
# the last element by torch.tensor_split may be empty
# e.g.
# torch.tensor_split(torch.tensor([1,2,3,4]), torch.tensor([2,4]))
# (tensor([1, 2]), tensor([3, 4]), tensor([], dtype=torch.int64))
break
best_seq_idx = processed_seqs + torch.argmax(tot_scores)

assert best_seq_idx < len(token_ids)
best_token_seqs = token_ids[best_seq_idx]
processed_seqs += tot_scores.nelement()
hyps.append(best_token_seqs)
scores.append(tot_scores.max().item())

assert len(hyps) == len(split_size)
else:
best_paths = one_best_decoding(lattices, use_double_scores=True)
scores = best_paths.get_tot_scores(
Expand Down