slahiriandClaude Opus 4.5
2a35976780
- Add modular core architecture with plug-and-play model support - Implement Florence-2 model with config-driven prompts and settings - Add platform layer for GPU detection, memory management, and state isolation - Support transformers >= 4.51.0 via kijai's bundled Florence2 implementation - Add caption cleaning and output formatting utilities - Update architecture docs with KISS principle design decisions 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
98 lines
2.4 KiB
Python
98 lines
2.4 KiB
Python
"""Output cleaning and formatting utilities."""
|
|
|
|
import re
|
|
|
|
|
|
def clean_caption(text: str) -> str:
|
|
"""
|
|
Clean and normalize caption output.
|
|
|
|
Removes:
|
|
- Thinking tags (<think>...</think>)
|
|
- Markdown formatting
|
|
- Common LLM prefixes ("Here's the caption:", etc.)
|
|
- Extra whitespace
|
|
|
|
Args:
|
|
text: Raw caption from model
|
|
|
|
Returns:
|
|
Cleaned caption string
|
|
"""
|
|
if not text:
|
|
return ""
|
|
|
|
# Remove thinking tags (common in some models)
|
|
text = re.sub(r"<think>.*?</think>", "", text, flags=re.DOTALL)
|
|
text = re.sub(r"<thinking>.*?</thinking>", "", text, flags=re.DOTALL)
|
|
|
|
# Remove markdown code blocks
|
|
text = re.sub(r"```[\s\S]*?```", "", text)
|
|
text = re.sub(r"`[^`]+`", "", text)
|
|
|
|
# Remove common LLM prefixes
|
|
prefixes = [
|
|
r"^Here['']?s?\s*(?:is\s*)?(?:the\s*)?(?:a\s*)?(?:caption|description|prompt)[:\s]*",
|
|
r"^(?:The\s*)?(?:image\s*)?(?:shows|depicts|features|contains)[:\s]*",
|
|
r"^(?:This\s*)?(?:image\s*)?(?:is\s*)?(?:a\s*)?(?:photo|picture|photograph)\s*of[:\s]*",
|
|
r"^(?:Caption|Description|Prompt)[:\s]*",
|
|
]
|
|
for prefix in prefixes:
|
|
text = re.sub(prefix, "", text, flags=re.IGNORECASE)
|
|
|
|
# Normalize whitespace
|
|
text = re.sub(r"\s+", " ", text)
|
|
text = text.strip()
|
|
|
|
return text
|
|
|
|
|
|
def truncate_to_words(text: str, max_words: int) -> str:
|
|
"""
|
|
Truncate text to a maximum number of words.
|
|
|
|
Attempts to end at a sentence boundary if possible.
|
|
|
|
Args:
|
|
text: Input text
|
|
max_words: Maximum word count
|
|
|
|
Returns:
|
|
Truncated text
|
|
"""
|
|
words = text.split()
|
|
|
|
if len(words) <= max_words:
|
|
return text
|
|
|
|
truncated = " ".join(words[:max_words])
|
|
|
|
# Try to end at sentence boundary
|
|
last_period = truncated.rfind(". ")
|
|
if last_period > len(truncated) * 0.7: # Only if near the end
|
|
return truncated[:last_period + 1]
|
|
|
|
return truncated + "..."
|
|
|
|
|
|
def format_as_tags(text: str) -> str:
|
|
"""
|
|
Format caption as comma-separated tags.
|
|
|
|
Args:
|
|
text: Caption text
|
|
|
|
Returns:
|
|
Comma-separated tag string
|
|
"""
|
|
# Already comma-separated
|
|
if "," in text and text.count(",") > 3:
|
|
return text
|
|
|
|
# Convert sentences to tags
|
|
# Remove periods and split on common delimiters
|
|
text = text.replace(".", ",")
|
|
text = re.sub(r"\s*,\s*", ", ", text)
|
|
|
|
return text.strip().rstrip(",")
|