Skip to content

fix(llama): grow the detokenize buffer when a token piece exceeds 32 bytes #2363

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
Anai-Guo wants to merge 1 commit into
base: main
Choose a base branch
from
+13 −0
Open

fix(llama): grow the detokenize buffer when a token piece exceeds 32 bytes #2363

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
13 changes: 13 additions & 0 deletions llama_cpp/_internals.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -196,6 +196,19 @@ def detokenize(self, tokens: List[int], special: bool = False) -> bytes:
n = llama_cpp.llama_token_to_piece(
self.vocab, llama_cpp.llama_token(token), buffer, size, 0, special
)
if n < 0:
# The piece did not fit; llama_token_to_piece wrote nothing and
# returned -(required size). Grow the buffer and ask again. The
# larger buffer is kept for the remaining tokens.
size = -n
buffer = (ctypes.c_char * size)()
n = llama_cpp.llama_token_to_piece(
self.vocab, llama_cpp.llama_token(token), buffer, size, 0, special
)
if n < 0:
raise RuntimeError(
f"Failed to detokenize: token={token} n={n} size={size}"
)
assert n <= size
output += bytes(buffer[:n])
# NOTE: Llama1 models automatically added a space at the start of the prompt
Expand Down
Loading