Skip to content

Gemma2 sampling crashes with AttributeError when Sampler builds end_tokens (BEGIN_OF_TOOL_RESPONSE) (#568) #569

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

Gemma2 sampling crashes with AttributeError when Sampler builds end_tokens (BEGIN_OF_TOOL_RESPONSE) (#568) #569

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
21 changes: 15 additions & 6 deletions gemma/gm/text/_sampler.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -341,15 +341,24 @@ def sample(

# TODO(epot): Donate the `init_state`, `last_state`

# Build end_tokens conditionally: only include BEGIN_OF_TOOL_RESPONSE if
# the tokenizer supports it (Gemma3+), not Gemma2.
base_end_tokens = (
self.tokenizer.special_tokens.EOS,
self.tokenizer.special_tokens.END_OF_TURN,
)
# Only add tool token if the tokenizer defines it (Gemma3+ only).
tool_token = getattr(
self.tokenizer.special_tokens, 'BEGIN_OF_TOOL_RESPONSE', None
)
if tool_token is not None:
base_end_tokens = (*base_end_tokens, tool_token)
end_tokens = (*base_end_tokens, *self._normalized_stop_tokens)

sampler = _sampler_loop.SamplerLoop(
# Static attributes. Changing those will trigger a recompilation.
model=self.model,
end_tokens=(
self.tokenizer.special_tokens.EOS,
self.tokenizer.special_tokens.END_OF_TURN,
self.tokenizer.special_tokens.BEGIN_OF_TOOL_RESPONSE,
*self._normalized_stop_tokens,
),
end_tokens=end_tokens, # ← Changed from tuple literal to variable
forbidden_tokens=self._normalized_forbidden_tokens,
sampling=sampling,
cache_length=self.cache_length,
Expand Down