65ce290104
llm decided to add additional safety layer to rag evaluation file save
idchlife2026-03-13 08:28:21 +03:00
236f44b2c3
evaluation for rag systems update, script for evaluating against questions
idchlife2026-03-13 08:20:18 +03:00
6c953a327f
evaluation for rag systems
idchlife2026-03-11 22:30:02 +03:00
5721bad117
fix for openai compatibility model with needed parameters
idchlife2026-03-09 10:21:39 +03:00
969d25209c
using openailike for llamaindex that will allow of usage of any openai compatible models via api, and not just hardcoded models inside default OpenAI class in llamaindex
idchlife2026-03-09 09:59:27 +03:00
6b3fa1cfaa
more sophisticated chat like retrieval for llamaindex
idchlife2026-02-26 19:02:05 +03:00
ba1b7abf0a
demo ui now has 3 tabs for each rag OSS solution
idchlife2026-02-26 15:05:18 +03:00
3e29ea70ed
langchain uploading new way of predefined paths from yandex disk
idchlife2026-02-26 00:01:47 +03:00
2c7ab06b3f
ragflow in the repository, with codex-created yandex disk plugin JUST IN CASE, also llamaindex enrichment with yandex disk predefined data
idchlife2026-02-25 11:28:29 +03:00
c29928cc89
prefect flow and task working inside langchain project
idchlife2026-02-16 16:04:47 +03:00
77c578c9e6
Prefect client prep for langchain
idchlife2026-02-16 15:12:44 +03:00
93d538ecc6
Checking properly source of the file for metadata, with instanceof
idchlife2026-02-11 16:23:27 +03:00
f5659675ec
- main feat: adaptation for async enrichment - added file_type, this will hold the "таблица", "презентация" and so on types - file source metadata is now taken either from local source or yandex disk.
idchlife2026-02-11 15:46:54 +03:00
7b52887558
Enrichment now processed via chunks. 2 documents -> into the vector storage. Also geussing source from the file extension
idchlife2026-02-11 11:23:50 +03:00
1e6ab247b9
Phase 12 done... loading via adaptive collection, yadisk or local
idchlife2026-02-10 22:19:27 +03:00
e9dd28ad55
Prep for Phase 12 of loading files for enrichment through the adaptive collections
idchlife2026-02-10 21:42:59 +03:00
06a3155b6b
Working Yandex Disk integration for loading files. Tests for local and Yandex
idchlife2026-02-10 20:42:07 +03:00
63c3e2c5c7
Adaptive Collection, and Phase 11 WIP
idchlife2026-02-10 20:12:43 +03:00
447ecaba39
enrichment with years, events
idchlife2026-02-10 13:20:19 +03:00
ce62fd50ed
Created this MD file to store things we need to look out to
idchlife2026-02-09 21:33:03 +03:00
2cb9b39bf2
removed test retrieval feature. off you go
idchlife2026-02-09 21:17:42 +03:00
f9c47c772f
llamaindex update + unpacking archives in data
idchlife2026-02-09 19:00:23 +03:00
ae8c00316e
Langchain plan phases for openai integration (openai compaible endpoint), server for retrieving data
idchlife2026-02-04 21:34:22 +03:00
ea4ce23cd9
Retrieval and also update on russian language
idchlife2026-02-04 16:51:50 +03:00
3dea3605ad
Enrichment for llamaindex. It goes for a long time using local model, so better use external model not local, for EMBEDDING
idchlife2026-02-04 16:06:01 +03:00
f36108d652
Vector storage Qdrant initialization and configuration
idchlife2026-02-04 01:10:07 +03:00
c37aec1d99
File extensions and libraries for llamaindex
idchlife2026-02-04 01:02:21 +03:00
fa26d77520
Cli with ping for llamaindex
idchlife2026-02-04 00:59:01 +03:00
86fd643e66
Start of work on Llamaindex framework
idchlife2026-02-04 00:49:45 +03:00
d354d3dcca
Working chat with AI agent with retrieving data
idchlife2026-02-04 00:02:53 +03:00
299ee0acb5
Working retrieval with the cli
idchlife2026-02-03 23:25:24 +03:00