worship / document_processing_agent.py

Commit History

Add name translation fix for OPUS-MT - correct 'sand geese' to 'Zhang Shaian'
e632f5c

Peter Yang commited on

Revert to OPUS-MT translation by default - better name handling
f04ca50

Peter Yang commited on

Fix translation hallucinations by adding stop sequences and improving cleanup
44c353a

Peter Yang commited on

Improve translation prompt to ensure completeness for titles and all sentences
531486f

Peter Yang commited on

Improve translation completeness: better paragraph splitting for headings, improved prompts, lower min length for titles
d3816fa

Peter Yang commited on

Fix: Enable PDF processing for content extraction and improve short translation handling
9124732

Peter Yang commited on

Fix Qwen2.5 translation truncation: increase max_length to 1024, dynamic max_new_tokens (300-800), less aggressive cleanup
1c3d743

Peter Yang commited on

Remove unused messages variable
52fab4c

Peter Yang commited on

Fix: Save bilingual file to current directory, prevent PDF/bilingual file duplication, ensure Message section only uses bilingual content
917635f

Peter Yang commited on

Integrate improved Qwen2.5 translation into document_processing_agent with fallback to OPUS-MT
b68a16e

Peter Yang commited on

Fix: Use actual bilingual file from document_sources instead of hardcoded test file
6cea639

Peter Yang commited on

Initial commit for HF Space deployment
3307396

Peter Yang commited on