Skip to content

fix text_offset of multi-token characters #1037

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

Merged
abetlen merged 2 commits into from
Dec 22, 2023
Merged

fix text_offset of multi-token characters #1037

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
68b69da
fix text_offsets for bytes tokens
twaka Dec 22, 2023
a583bf2
fix
twaka Dec 22, 2023
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
18 changes: 13 additions & 5 deletions llama_cpp/llama.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1551,7 +1551,9 @@ def logit_bias_processor(
"utf-8", errors="ignore"
)
text_offset = len(prompt) + len(
self.detokenize(completion_tokens[:returned_tokens])
self.detokenize(completion_tokens[:returned_tokens]).decode(
"utf-8", errors="ignore"
)
)
token_offset = len(prompt_tokens) + returned_tokens
logits = self._scores[token_offset - 1, :]
Expand Down Expand Up @@ -1789,13 +1791,19 @@ def logit_bias_processor(
]
all_logprobs = Llama.logits_to_logprobs(self._scores)[token_offset:]
# TODO: may be able to change this loop to use np.take_along_dim
for token, token_str, logprobs_token in zip(
all_tokens, all_token_strs, all_logprobs
for idx, (token, token_str, logprobs_token) in enumerate(
zip(all_tokens, all_token_strs, all_logprobs)
):
if token == self.token_bos():
continue
text_offsets.append(text_offset)
text_offset += len(token_str)
text_offsets.append(
text_offset
+ len(
self.detokenize(all_tokens[:idx]).decode(
"utf-8", errors="ignore"
)
)
)
tokens.append(token_str)
sorted_logprobs = list(
sorted(
Expand Down