Skip to content

fix: eliminate hard-coded vocab definitions to make the Whisper model compatible with custom vocabularies and embedding layer lengths #3555

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
Jaffe2718 wants to merge 4 commits into
base: master
Choose a base branch
from
+15 −16
Open

fix: eliminate hard-coded vocab definitions to make the Whisper model compatible with custom vocabularies and embedding layer lengths #3555

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
79b0c01
Avoid hard-coding definition vocabulary to be compatible with differe…
Jaffe2718 Dec 4, 2025
43d873c
modify comment
Jaffe2718 Dec 4, 2025
ccd9b6e
fix num_language(): incorrect after loading special tokens
Jaffe2718 Dec 6, 2025
1ab1804
fix convert script: remove special token `<|endoftext|>`
Jaffe2718 Dec 6, 2025
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
2 changes: 2 additions & 0 deletions models/convert-h5-to-ggml.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -107,6 +107,8 @@ def bytes_to_unicode():
fname_out = dir_out / "ggml-model.bin"

tokens = json.load(open(dir_tokenizer / "vocab.json", "r", encoding="utf8"))
if "<|endoftext|>" in tokens:
del tokens["<|endoftext|>"]
Comment on lines +110 to +111
Copy link
Author

@Jaffe2718 Jaffe2718 Dec 6, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

When I use ggml-tiny.en.bin recognition, the result is also empty, the same reason why the tests fail

Copy link
Author

@Jaffe2718 Jaffe2718 Dec 6, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

If the last commit was Migrate from HG dataset into HG model, it is necessary that these models need to be reconverted if they were last generated with this script, otherwise, <|endoftext|> will be written into common tokens.


# use 16-bit or 32-bit floats
use_f16 = True
Expand Down
29 changes: 13 additions & 16 deletions src/whisper.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -453,7 +453,7 @@ struct whisper_vocab {
}

int num_languages() const {
return n_vocab - 51765 - (is_multilingual() ? 1 : 0);
return token_translate - token_sot - 1;
}
};

Expand Down Expand Up @@ -1621,22 +1621,19 @@ static bool whisper_model_load(struct whisper_model_loader * loader, whisper_con
//printf("%s: vocab[%d] = '%s'\n", __func__, i, word.c_str());
}

vocab.n_vocab = model.hparams.n_vocab;
if (vocab.is_multilingual()) {
vocab.token_eot++;
vocab.token_sot++;
vocab.n_vocab = model.hparams.n_vocab; // all tokens, including special tokens

// account for variable number of language tokens
const int dt = vocab.num_languages() - 98;

vocab.token_translate += dt;
vocab.token_transcribe += dt;
vocab.token_solm += dt;
vocab.token_prev += dt;
vocab.token_nosp += dt;
vocab.token_not += dt;
vocab.token_beg += dt;
}
vocab.token_eot = n_vocab; // <|endoftext|> 50256 for en, 50257 for multilingual, others for custom model
vocab.token_sot = n_vocab + 1; // <|startoftext|>
// [n_vocab + 2, vocab.n_vocab - 1507) are language tokens
// num_language = vocab.token_translate - vocab.token_sot - 1 = vocab.n_vocab - n_vocab - 1509
vocab.token_translate = vocab.n_vocab - 1507; // <|translate|>
vocab.token_transcribe = vocab.n_vocab - 1506; // <|transcribe|>
vocab.token_solm = vocab.n_vocab - 1505; // <|startoflm|>
vocab.token_prev = vocab.n_vocab - 1504; // <|startofprev|>
vocab.token_nosp = vocab.n_vocab - 1503; // <|nospeech|>
vocab.token_not = vocab.n_vocab - 1502; // <|notimestamps|>
vocab.token_beg = vocab.n_vocab - 1501; // timestamps from <|0.00|> to <|30.00|>, 1501 tokens

if (n_vocab < model.hparams.n_vocab) {
WHISPER_LOG_INFO("%s: adding %d extra tokens\n", __func__, model.hparams.n_vocab - n_vocab);
Expand Down