Commit graph

96 commits

Author SHA1 Message Date
Yiorgis Gozadinos
390397570a
Migrate QA agent to Pydantic AI. 2025-08-15 21:13:17 +02:00
Yiorgis Gozadinos
bf8a4f22be
Make radius a parameter we can pass to expand_context() 2025-08-12 16:53:45 +02:00
Yiorgis Gozadinos
99f7401817
Merge expanded context if possible 2025-08-12 15:02:07 +02:00
Yiorgis Gozadinos
e7c2a105ed
Use the expand_context when searching in the qa agent 2025-08-12 14:51:33 +02:00
Yiorgis Gozadinos
eb13426aae
expand_context() by using the adjacent chunks 2025-08-12 14:14:18 +02:00
Yiorgis Gozadinos
2f42b1a3ac
Get adjacent chunks method 2025-08-12 13:24:04 +02:00
Yiorgis Gozadinos
842e4a882d
Make citations optional, through cite parameter 2025-08-12 13:19:11 +02:00
Yiorgis Gozadinos
a0d0359618
Change prompt and tool definition to to include citations in the QA agent. 2025-08-12 13:05:37 +02:00
Yiorgis Gozadinos
6f50706e8a
Give the results of search to the qa agent as json including the document uri 2025-08-12 12:40:55 +02:00
Yiorgis Gozadinos
e39be89666
Merge pull request #33 from ggozad/feat/ollama-reranker
Add custom reranker based on Ollama.
2025-08-11 11:18:42 +02:00
Yiorgis Gozadinos
022f8677b3
Merge pull request #32 from ggozad/chore/slim-deps
Make mxbai reranker optional. Only use a reranker if mxbai or cohere are installed.
2025-08-11 11:18:27 +02:00
Yiorgis Gozadinos
9648cb97cb
Fix URL handling in add-src 2025-08-11 11:10:29 +02:00
Yiorgis Gozadinos
ffe867f7bb
Allow to disable reranking by having an empty provider 2025-08-10 21:02:17 +02:00
Yiorgis Gozadinos
5103e6c2d7
Add Ollama reranker 2025-08-10 20:50:17 +02:00
Yiorgis Gozadinos
8d98eed2e2
Make mxbai reranker optional. Only use a reranker if mxbai or cohere are installed 2025-08-10 19:49:14 +02:00
Yiorgis Gozadinos
0b3477c46c
Rebuild now re-adds files. 2025-08-08 18:30:53 +02:00
Yiorgis Gozadinos
22ed9ca51e
Disable python warning unless in development 2025-08-08 17:31:42 +02:00
Yiorgis Gozadinos
ad4b655a31
Operate on DoclingDocument instead of Document 2025-08-08 17:24:34 +02:00
Yiorgis Gozadinos
6a3c85e6ae
text_to_docling_document() util 2025-08-08 15:38:57 +02:00
Yiorgis Gozadinos
b92f25d6fb
Make FileReader return a DoclingDocument instead of a md string 2025-08-08 14:45:24 +02:00
Yiorgis Gozadinos
7906b1b026
Use absolute paths when add-src 2025-08-08 14:34:13 +02:00
Yiorgis Gozadinos
af95b5f3eb
Replace loop.run_until_complete() with asyncio.run() 2025-08-08 12:41:50 +02:00
Yiorgis Gozadinos
9b4942401f
Fix file extension typo 2025-08-01 17:15:14 +02:00
Yiorgis Gozadinos
ec1bc21a91
Contextualize chunks 2025-08-01 16:54:51 +02:00
Yiorgis Gozadinos
a453853211
Keep using the openai tokenizer 2025-08-01 16:36:24 +02:00
Yiorgis Gozadinos
91f6122ef7
Basic chunking using docling 2025-08-01 16:27:44 +02:00
Yiorgis Gozadinos
75c1cdb36c
Merge pull request #26 from ggozad/feat/docling
Use docling for parsing documents
2025-08-01 16:10:34 +02:00
Yiorgis Gozadinos
4bfd54b2ec
Merge pull request #25 from ggozad/feat/sqlite_wal
Use Write-Ahead Logging jorunaling mode for SQLite concurrency
2025-08-01 16:10:13 +02:00
Yiorgis Gozadinos
14d8b18f1d
Wrap code files in md code blocks 2025-08-01 10:13:46 +02:00
Yiorgis Gozadinos
5a08816ed6
Replace markitdown with docling 2025-08-01 09:59:53 +02:00
Yiorgis Gozadinos
0f2809cb8d
Use Write-Ahead Logging jorunaling mode for SQLite concurrency 2025-08-01 09:07:31 +02:00
Yiorgis Gozadinos
8052442052
Add -v --version option to display version 2025-08-01 09:00:20 +02:00
Yiorgis Gozadinos
74eea9284a
create_document now optionally accepts pre-computed chunks 2025-07-26 10:42:27 +03:00
Yiorgis Gozadinos
6a7d2ff319
Merge pull request #20 from ggozad/fix/honour-default_data_dir
Honour DEFAULT_DATA_DIR env variable in CLI
2025-07-23 13:52:25 +03:00
Yiorgis Gozadinos
72ce212ad4
Honour DEFAULT_DATA_DIR env variable 2025-07-23 13:49:27 +03:00
Yiorgis Gozadinos
dfe23482da
Tune parameters 2025-07-20 22:45:24 +03:00
Yiorgis Gozadinos
ad9879802a
Fix incorrect syntax for uv pip install with extras 2025-07-20 22:40:12 +03:00
Yiorgis Gozadinos
942be2271c
Update docs 2025-07-20 11:33:09 +03:00
Yiorgis Gozadinos
c1d5334804
Update benchmarks with reranking 2025-07-19 19:02:47 +03:00
Yiorgis Gozadinos
2347612fb6
Optionally perform reranking when searching 2025-07-19 13:12:34 +03:00
Yiorgis Gozadinos
1437942edc
Make reranker a global module object to avoid re-initialization 2025-07-19 12:51:22 +03:00
Yiorgis Gozadinos
2343ab7751
Return tuple (Chunk, score,) in rerank() 2025-07-19 12:16:17 +03:00
Yiorgis Gozadinos
f25416b0ff
Properly test cohere reranker 2025-07-18 20:55:28 +03:00
Yiorgis Gozadinos
044e1b0a06
mixedbread reranker as default 2025-07-18 20:29:24 +03:00
Yiorgis Gozadinos
0f8124fd98
Simply embedders 2025-07-17 13:17:47 +03:00
Yiorgis Gozadinos
859860318a
Cohere reranker 2025-07-17 13:17:16 +03:00
Yiorgis Gozadinos
20beb23a38
Fix docstrings 2025-07-16 19:06:54 +03:00
Yiorgis Gozadinos
4d8088efa8
Log db upgrades 2025-07-16 18:32:58 +03:00
Yiorgis Gozadinos
0bf5606117
Store haiku.rag version, enabling upgrades 2025-07-16 18:21:41 +03:00
Yiorgis Gozadinos
f4aef4eb9c
Check pypi version on startup 2025-07-14 12:27:28 +03:00