q
|
e468c4768b
|
Migrate to multi-file architecture: smarter chunking + fixed RERANK_LIMIT
index: message-based windowed chunking (5 msgs/1h gap), better unicode
cleaning, separate dense (with timestamps)/sparse content renderers,
BM25 preload on startup, ThreadPoolExecutor(4), UVICORN_WORKERS=4,
Dockerfile copies all *.py
search: proper multi-module structure (query_builder, retrieval, rerank,
aggregation), RERANK_LIMIT 60→15 (fixes 429 errors), extra dense vectors
for variants/hyde, date+asker metadata filters, httpx pool (100/20/30s),
BM25 preload on startup, Dockerfile copies all *.py
68/68 unit tests passing
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
|
2026-04-18 21:49:14 +03:00 |
|
q
|
c9083c285f
|
Optimize for 4 cores: BM25 preload, httpx pool, orjson, fix lambda
- index: UVICORN_WORKERS 8→4, lifespan BM25 preload, explicit ThreadPoolExecutor(4), orjson
- search: lifespan BM25 preload, httpx limits (max_conn=100, keepalive=20, timeout=30s), fix asyncio.to_thread lambda, orjson
- both: ORJSONResponse as default_response_class
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
|
2026-04-18 21:21:28 +03:00 |
|