Yiorgis Gozadinos
|
98e0124ee0
|
Use 3.13 for development
|
2025-10-09 10:18:10 +03:00 |
|
Yiorgis Gozadinos
|
f3dfae2670
|
Merge pull request #94 from ggozad/feat/mcp-registry
Register haiku.rag with the MCP server registry
|
2025-10-08 11:16:29 +03:00 |
|
Yiorgis Gozadinos
|
c9fc0cd750
|
MCP server registry
|
2025-10-08 11:10:54 +03:00 |
|
Yiorgis Gozadinos
|
09528221b5
|
Merge pull request #93 from ggozad/feat/mcp-enhancements
Add ask, deep ask, research as MCP tools.
|
2025-10-08 09:22:04 +03:00 |
|
Yiorgis Gozadinos
|
43dac018c0
|
Test mcp tools
|
2025-10-08 09:20:26 +03:00 |
|
Yiorgis Gozadinos
|
e7aca6292a
|
Add ask, research as tools to mcp
|
2025-10-08 09:13:21 +03:00 |
|
Yiorgis Gozadinos
|
8323cbedc3
|
vb
|
2025-10-07 14:08:34 +03:00 |
|
Yiorgis Gozadinos
|
84b0221495
|
Merge pull request #92 from ggozad/feat/customizable-system-prompt
Customizable system prompt for Q/A agent
|
2025-10-07 14:07:34 +03:00 |
|
Yiorgis Gozadinos
|
0991df6d9e
|
Make it possible to override the QA agent system prompt using the client
|
2025-10-07 13:15:58 +03:00 |
|
Yiorgis Gozadinos
|
496f2777cb
|
Use custom prompt when evaluating with the Wix dataset to make the QA agent act as an assistant
|
2025-10-07 13:12:31 +03:00 |
|
Yiorgis Gozadinos
|
d87cafdea4
|
Merge pull request #91 from ggozad/feat/auto-vacuming
Auto-vacuuming versions older than VACUUM_RETENTION_SECONDS (default=60sec)
|
2025-10-07 12:34:32 +03:00 |
|
Yiorgis Gozadinos
|
95297432eb
|
Doc updates
|
2025-10-07 12:33:14 +03:00 |
|
Yiorgis Gozadinos
|
5e4928ccad
|
Make sure vacuum task has been complete before exiting the client
|
2025-10-07 12:27:55 +03:00 |
|
Yiorgis Gozadinos
|
094f0e0d44
|
Centralize vacuuming
|
2025-10-07 12:05:25 +03:00 |
|
Yiorgis Gozadinos
|
ce11efd17c
|
Auto-vacuum when inserting/updating
|
2025-10-07 10:54:56 +03:00 |
|
Yiorgis Gozadinos
|
c6eaf9d5cd
|
Add vacuum retention setting
|
2025-10-07 10:47:28 +03:00 |
|
Yiorgis Gozadinos
|
4b4af30e56
|
Update docling & lance
|
2025-10-07 09:02:21 +03:00 |
|
Yiorgis Gozadinos
|
54a22a6f98
|
Merge branch 'feat/success-metrics'
|
2025-10-06 14:55:40 +03:00 |
|
Yiorgis Gozadinos
|
6f181e07ff
|
Add Success@K metrics to retrieval benchmarks
|
2025-10-06 14:54:17 +03:00 |
|
Yiorgis Gozadinos
|
cf2af489b5
|
Allow QA agent to retry
|
2025-10-06 11:21:42 +03:00 |
|
Yiorgis Gozadinos
|
99463cc39b
|
vb
|
2025-10-01 08:41:51 +03:00 |
|
Yiorgis Gozadinos
|
39f804e035
|
Merge pull request #89 from ggozad/feat/deep-qa
Deep Question/Answer graph agent.
|
2025-10-01 08:41:07 +03:00 |
|
Yiorgis Gozadinos
|
a6275fcf25
|
Rename nodes to make it clear they are part of Q/A
|
2025-09-30 16:59:36 +03:00 |
|
Yiorgis Gozadinos
|
49eac86259
|
Add verbose flag to ask
|
2025-09-30 16:59:36 +03:00 |
|
Yiorgis Gozadinos
|
2dfb1882dd
|
Deep QA multi-agent
|
2025-09-30 16:59:36 +03:00 |
|
Yiorgis Gozadinos
|
42c54a8d25
|
Shuffle reusable nodes under graph/
|
2025-09-30 16:59:36 +03:00 |
|
Yiorgis Gozadinos
|
fa698cca28
|
vb
|
2025-09-30 16:58:52 +03:00 |
|
Yiorgis Gozadinos
|
999f9408f6
|
Merge pull request #90 from ggozad/fix/rebuild-missing-docs
Fix rebuilding for missing documents by re-embedding.
|
2025-09-30 16:57:58 +03:00 |
|
Yiorgis Gozadinos
|
05c382fcb5
|
When rebuilding if a document does not exist anymore, re-embed the chunks instead of failing
|
2025-09-30 16:57:02 +03:00 |
|
Yiorgis Gozadinos
|
43b9cd50ba
|
Move evaluations to src/
|
2025-09-30 13:21:53 +03:00 |
|
Yiorgis Gozadinos
|
6bcc251e5d
|
Merge pull request #88 from ggozad/fix/clarify-incompatible-settings
Make exception for incompatible settings more user-friendly
|
2025-09-30 12:22:37 +03:00 |
|
Yiorgis Gozadinos
|
9318ffc302
|
Make exception for incompatible settings more user-friendly
|
2025-09-30 12:19:08 +03:00 |
|
Yiorgis Gozadinos
|
6a1d7b3fa9
|
Ignore tests/data where we store the qa corpus
|
2025-09-30 11:42:28 +03:00 |
|
Yiorgis Gozadinos
|
2ae11200b4
|
Merge pull request #87 from ggozad/feat/multiple-eval-datasets
Run evaluations with alternate datasets. Use WixQA as an example.
|
2025-09-30 11:32:40 +03:00 |
|
Yiorgis Gozadinos
|
06dbdbd474
|
Update docs for wix benchmarks
|
2025-09-30 11:31:52 +03:00 |
|
Yiorgis Gozadinos
|
abfc796c42
|
Use gpt-oss for evaluation LLMJudge, allow it to retry if it fails
|
2025-09-30 11:31:51 +03:00 |
|
Yiorgis Gozadinos
|
f62d3e552b
|
Exclude evaluations, test, docs, github actions from builds
|
2025-09-30 11:31:51 +03:00 |
|
Yiorgis Gozadinos
|
6f20d9fd6d
|
Update recall of qwen3 embeddings with mxbai reranking
|
2025-09-30 11:31:51 +03:00 |
|
Yiorgis Gozadinos
|
9a859c6ee5
|
Add option to skip db in evals
|
2025-09-30 11:31:51 +03:00 |
|
Yiorgis Gozadinos
|
842e166041
|
Adapt how we measure recall when using datasets with multiple sources
|
2025-09-30 11:31:51 +03:00 |
|
Yiorgis Gozadinos
|
eb2fc67e1c
|
When populating the eval db, check if chunks for the document have been created. Protects against interrupts
|
2025-09-30 11:31:50 +03:00 |
|
Yiorgis Gozadinos
|
55dc4bbdd2
|
Return an empty array when request en empty embedding
|
2025-09-30 11:31:50 +03:00 |
|
Yiorgis Gozadinos
|
57e86d00fd
|
Refactor evaluations so that we can perform with multiple datasets..
Introduce Wix dataset.
|
2025-09-30 11:31:50 +03:00 |
|
Yiorgis Gozadinos
|
73209b203c
|
Merge pull request #86 from ggozad/feat/pydantic-evaluations
Use pydantic evals for running our own evaluations
|
2025-09-30 11:31:20 +03:00 |
|
Yiorgis Gozadinos
|
19ad1b9d7d
|
improve output
|
2025-09-30 11:30:35 +03:00 |
|
Yiorgis Gozadinos
|
87f28ed96b
|
Update docs
|
2025-09-30 11:30:35 +03:00 |
|
Yiorgis Gozadinos
|
7cc4e13b57
|
Make benchmarks use pydantic evals
|
2025-09-30 11:30:34 +03:00 |
|
Yiorgis Gozadinos
|
53557b0629
|
Merge pull request #85 from ggozad/chore/qwen3-embeddings
Change default embedding to qwen3-embedding. Update benchmarks.
|
2025-09-30 11:30:02 +03:00 |
|
Yiorgis Gozadinos
|
14168db795
|
Fix embedding client test
|
2025-09-26 13:52:06 +03:00 |
|
Yiorgis Gozadinos
|
f0333e33ac
|
Change default embedding to qwen3-embedding. Update benchmarks using qwen3-embedding and gpt-oss
|
2025-09-26 13:23:38 +03:00 |
|