Skip to content

Allow no-timestamp Whisper decoding to use full context #2075

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

Open
Emre-Akgul wants to merge 1 commit into
base: master
Choose a base branch
from
Open

Allow no-timestamp Whisper decoding to use full context #2075

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.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 6 additions & 2 deletions src/models/whisper.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -292,6 +292,7 @@ namespace ctranslate2 {
}

const dim_t total_max_length = options.max_length;
const bool without_timestamps = prompts[0][prompt_length - 1] == _no_timestamps_id;

DecodingOptions decoding_options;
decoding_options.start_step = start_step;
Expand All @@ -300,7 +301,10 @@ namespace ctranslate2 {
decoding_options.length_penalty = options.length_penalty;
decoding_options.repetition_penalty = options.repetition_penalty;
decoding_options.no_repeat_ngram_size = options.no_repeat_ngram_size;
decoding_options.max_length = std::min(total_max_length / 2, total_max_length - start_step);
decoding_options.max_length = without_timestamps
? total_max_length - start_step
: std::min(total_max_length / 2,
total_max_length - start_step);
decoding_options.sampling_topk = options.sampling_topk;
decoding_options.sampling_temperature = options.sampling_temperature;
decoding_options.num_hypotheses = options.num_hypotheses;
Expand Down Expand Up @@ -329,7 +333,7 @@ namespace ctranslate2 {
decoding_options.logits_processors.emplace_back(no_speech_probs_processor);
}

if (prompts[0][prompt_length - 1] != _no_timestamps_id) {
if (!without_timestamps) {
const size_t timestamp_begin_id = _no_timestamps_id + 1;
const size_t timestamp_end_id = vocabulary.size() - 1;
const size_t max_initial_timestamp_id = timestamp_begin_id + options.max_initial_timestamp_index;
Expand Down
Loading