Skip to content
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

fix: When the thinking process label is empty, do not process the thinking process #2179

Merged
merged 1 commit into from
Feb 9, 2025
Merged

fix: When the thinking process label is empty, do not process the thinking process #2179

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
15 changes: 12 additions & 3 deletions apps/application/flow/tools.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -23,14 +23,23 @@ def __init__(self, reasoning_content_start, reasoning_content_end):
self.all_content = ""
self.reasoning_content_start_tag = reasoning_content_start
self.reasoning_content_end_tag = reasoning_content_end
self.reasoning_content_start_tag_len = len(reasoning_content_start)
self.reasoning_content_end_tag_len = len(reasoning_content_end)
self.reasoning_content_end_tag_prefix = reasoning_content_end[0]
self.reasoning_content_start_tag_len = len(
reasoning_content_start) if reasoning_content_start is not None else 0
self.reasoning_content_end_tag_len = len(reasoning_content_end) if reasoning_content_end is not None else 0
self.reasoning_content_end_tag_prefix = reasoning_content_end[
0] if self.reasoning_content_end_tag_len > 0 else ''
self.reasoning_content_is_start = False
self.reasoning_content_is_end = False
self.reasoning_content_chunk = ""

def get_reasoning_content(self, chunk):
# 如果没有开始思考过程标签那么就全是结果
if self.reasoning_content_start_tag is None or len(self.reasoning_content_start_tag) == 0:
self.content += chunk.content
return {'content': chunk.content, 'reasoning_content': ''}
# 如果没有结束思考过程标签那么就全部是思考过程
if self.reasoning_content_end_tag is None or len(self.reasoning_content_end_tag) == 0:
return {'content': '', 'reasoning_content': chunk.content}
self.all_content += chunk.content
if not self.reasoning_content_is_start and len(self.all_content) >= self.reasoning_content_start_tag_len:
if self.all_content.startswith(self.reasoning_content_start_tag):
Copy link
Contributor Author

Choose a reason for hiding this comment

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

The provided Python code includes some minor improvements, such as handling cases where either start or end tags might be missing by setting appropriate attributes to zero or an empty string when necessary. However, there is room for further optimization and clarity:

  1. Simplify Logic: The logic for determining start/end of the reasoning content can be simplified by checking conditions once and directly returning results based on these checks.

  2. Avoid Redundant Checks: The redundant len() calls for both tags within the same condition can be removed.

  3. Use Default Parameters: Consider using default parameters in method signatures that accept optional arguments with meaningful names, which improves readability and maintainability.

Here's an improved version of the code:

def __init__(self, reasoning_content_start="", reasoning_content_end=""):
        self.all_content = ""
        self.reasoning_content_start_tag = reasoning_content_start
        self.reasoning_content_end_tag = reasoning_content_end

        self.start_prefix_length = len(reasoning_content_start) if reasoning_content_start else 0
        self.end_suffix_length = len(reasoning_content_end) if reasoning_content_end else 0
        self.end_tag_prefix = reasoning_content_end[0] if self.end_suffix_length > 0 else ''

    def get_reasoning_content(self, chunk):
        all_content = self.all_content + chunk.content

        # Start of reasoning content detected
        if all_content.startswith(self.reasoning_content_start_tag):
            result = {
                'content': '',
                'reasoning_content': ''.join(all_content.split(chunk.content)[1:])
            }
            return result

        # End of reasoning content detected
        elif all_content.endswith(self.reasoning_content_end_tag):
            reason_text = "".join(all_content.split(chunk.content))
            return {
                'content': chunk.content,
                'reasoning_content': reason_text[len(self.reasoning_content_end_tag):].strip()
            }

        # No tag matches (either at the beginning or the middle)
        else:
            self.all_content = ""
            return {'content': chunk.content, 'reasoning_content': ''}

Key Improvements:

  • Single Check for Tags: The code now uses only one check (startswith followed by endswith) instead of two separate checks, reducing redundancy.
  • Default Parameter Names: Added descriptive parameter names (reasoning_content_start, reasoning_content_end) to improve understanding of their purpose.
  • String Manipulation Simplified: Used .split() and slicing to simplify extracting text between tags.
  • Return Statement Clarity: Each branch of the logic returns a dictionary with clear keys indicating whether it contains main content or reasoning content.

Expand Down