mirror of
https://github.com/agent0ai/agent-zero.git
synced 2026-06-02 07:11:56 +00:00
b2ead06a4e
Run LiteParse in a subprocess so native parser crashes cannot take down the Web UI process. Bound parser concurrency and LiteParse workers for multi-chat stability, seed Q&A context with leading document chunks for title/abstract grounding, and keep a small-document fallback when vector search returns no chunks.
34 lines
1.4 KiB
YAML
34 lines
1.4 KiB
YAML
# Document Query Plugin Configuration
|
|
# All timeout values in seconds
|
|
|
|
# --- Timeouts ---
|
|
fetch_timeout: 30 # HTTP fetch connect/read timeout
|
|
fetch_retries: 3 # HTTP retry attempts
|
|
fetch_retry_backoff: 1.0 # delay between HTTP retry attempts
|
|
per_document_timeout: 60 # max time for a single document parse
|
|
gather_timeout: 120 # max time for all documents combined in one call
|
|
|
|
# --- Parser settings ---
|
|
parser_concurrency: 1 # max parser jobs running across all chats in this process
|
|
context_intro_chunks: 2 # always include leading chunks per document for title/abstract grounding
|
|
chunk_size: 1000
|
|
chunk_overlap: 100
|
|
search_threshold: 0.5
|
|
search_limit: 100
|
|
max_remote_bytes: 52428800 # 50 MB
|
|
|
|
# --- Feature flags ---
|
|
liteparse_enabled: true # prefer LiteParse before legacy parser fallbacks
|
|
liteparse_ocr_enabled: true
|
|
liteparse_ocr_language: eng
|
|
liteparse_ocr_server_url:
|
|
liteparse_tessdata_path:
|
|
liteparse_max_pages: 1000
|
|
liteparse_target_pages:
|
|
liteparse_dpi: 150
|
|
liteparse_preserve_very_small_text: false
|
|
liteparse_output_format: text
|
|
liteparse_num_workers: 1 # LiteParse defaults to CPU cores - 1; cap it for web runtime stability
|
|
liteparse_subprocess: true # isolate LiteParse native runtime crashes from the Web UI process
|
|
pdf_ocr_fallback: true # enable legacy Tesseract fallback after PyMuPDF
|
|
thread_offload: true # offload sync parsers to thread pool
|