8000 Programmatically remove control characters from LLM response by xtzie · Pull Request #1367 · 567-labs/instructor · GitHub
[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
Skip to content

Programmatically remove control characters from LLM response #1367

New issue

Have a question ab 8000 out 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
merged 1 commit into from
Mar 3, 2025
Merged

Programmatically remove control characters from LLM response #1367

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
31 changes: 27 additions & 4 deletions instructor/dsl/partial.py
View file
Open in desktop
< 8535 tr class="inline-comments js-inline-comments-container js-skip-tagsearch">
Original file line number Diff line number Diff line change
Expand Up @@ -29,6 +29,7 @@

from instructor.mode import Mode
from instructor.utils import extract_json_from_stream, extract_json_from_stream_async
import re

T_Model = TypeVar("T_Model", bound=BaseModel)

Expand All @@ -47,6 +48,18 @@ class PartialLiteralMixin:
pass


def remove_control_chars(s):
return re.sub(r"[\x00-\x1F\x7F-\x9F]", "", s)


def process_potential_object(potential_object, partial_mode, partial_model, **kwargs):
obj = from_json(
(potential_object.strip() or "{}").encode(), partial_mode=partial_mode
)
obj = partial_model.model_validate(obj, strict=None, **kwargs)
return obj


def _process_generic_arg(
arg: Any,
make_fields_optional: bool = False,
Expand Down Expand Up @@ -213,12 +226,22 @@ def model_from_chunks(
partial_mode = (
"on" if issubclass(cls, PartialLiteralMixin) else "trailing-strings"
)
chunk_buffer = []
for chunk in json_chunks:
potential_object += chunk
obj = from_json(
(potential_object.strip() or "{}").encode(), partial_mode=partial_mode
chunk_buffer += chunk
Copy link
Contributor

Choose a reason for hiding this comment

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

Using += on a list with a string will extend the list by characters. Replace with append() to keep chunks intact.

Suggested change
chunk_buffer += chunk
chunk_buffer.append(chunk)

if len(chunk_buffer) < 2:
continue
potential_object += remove_control_chars("".join(chunk_buffer))
chunk_buffer = []
obj = process_potential_object(
potential_object, partial_mode, partial_model, **kwargs
)
yield obj
if chunk_buffer:
potential_object += remove_control_chars(chunk_buffer[0])
Copy link
Contributor

Choose a reason for hiding this comment

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

Potential bug: In the final check, only processing chunk_buffer[0] ignores remaining buffered content. Consider joining all elements (e.g., using ''.join(chunk_buffer)).

Suggested change
potential_object += remove_control_chars(chunk_buffer[0])
potential_object += remove_control_chars(''.join(chunk_buffer))

obj = process_potential_object(
potential_object, partial_mode, partial_model, **kwargs
)
obj = partial_model.model_validate(obj, strict=None, **kwargs)
yield obj

@classmethod
Expand Down
0