Skip to content

fix: add missing pre-tokenizer type for GPT-2 BPE models #508

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
raphaelbgr wants to merge 1 commit into
base: main
Choose a base branch
from
Open

fix: add missing pre-tokenizer type for GPT-2 BPE models #508

Show file tree
Hide file tree
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
2 changes: 1 addition & 1 deletion utils/convert-hf-to-gguf-bitnet.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -957,7 +957,7 @@ class BitnetModel(Model):
model_arch = gguf.MODEL_ARCH.BITNET

def set_vocab(self):
self._set_vocab_sentencepiece()
self._set_vocab_gpt2()

def set_gguf_parameters(self):
super().set_gguf_parameters()
Expand Down
3 changes: 3 additions & 0 deletions utils/convert-ms-to-gguf-bitnet.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1208,6 +1208,9 @@ def extract_vocabulary_from_model(self, vocab: Vocab) -> tuple[list[bytes], list
def add_meta_vocab(self, vocab: Vocab) -> None:
# Ensure that tokenizer_model is added to the GGUF model
self.gguf.add_tokenizer_model(vocab.tokenizer_model)
# Add pre-tokenizer type for BPE models (required for correct tokenization)
if vocab.tokenizer_model == "gpt2":
self.gguf.add_token_pre_type("gpt-2")
# Extract model vocabulary for model conversion
tokens, scores, toktypes = self.extract_vocabulary_from_model(vocab)

Expand Down