Commit Graph

26 Commits

Author SHA1 Message Date
Alex 8a5e1e8d98 cleanups 2023-02-15 20:29:09 +00:00
Alex 4d1ff8238d switching between llms 2023-02-15 18:40:23 +00:00
Alex f9fe3f2f48 Merge branch 'main' into custom-llm 2023-02-15 14:42:57 +00:00
EricGao888 aeac186484 Add retry strategy to increase stability 2023-02-15 17:29:39 +08:00
Pavel d57c7b0296 -y-description 2023-02-15 13:10:30 +04:00
冯不游 b83589a308 feat: add support for directory list
example: `python ingest.py --dir inputs1 --dir another --dir ../inputs`,
the outputs will be in `outputs/input_folder_name/`
2023-02-15 02:30:39 +08:00
Alex 5883ce2685 Merge pull request #87 from arc53/ingest-cli
Ingest cli
2023-02-14 16:20:05 +00:00
Pavel af20c7298a new-ingest
Ingest with a CLI
2023-02-14 19:37:07 +04:00
冯不游 636783ca8a fix: avoid second error issue 2023-02-14 22:29:17 +08:00
冯不游 458f2a3ff3 fix: restore index back when continue process 2023-02-14 22:05:16 +08:00
Alex 046fbebf56 Enable other llm's 2023-02-14 13:06:28 +00:00
冯不游 3ab02ca959 feat: compatible with markdown 2023-02-13 11:00:34 +08:00
Alex e88ff885fe Merge pull request #75 from arc53/rst-interpreters 2023-02-12 18:32:20 +00:00
Pavel b1a6ebffba Directives + Interpreted
Some additional filters for rst parsing
2023-02-12 22:29:40 +04:00
Alex 205be538a3 fix dbqa, with new chain type, also fix for doc export 2023-02-12 17:58:54 +00:00
Alex 9228005a7e chunked embedding 2023-02-12 16:25:01 +00:00
vintro 2a203aa547 Create __init__.py
otherwise running `python ingest.py` complains about `parser` not being a package
2023-02-10 19:49:00 -05:00
Alex 37ad3b35c8 Update ingest.py 2023-02-10 17:30:37 +00:00
Alex d642782a5a move folder 2023-02-10 16:10:53 +00:00
Pavel 8c4fcff617 requirement 2023-02-10 19:47:51 +04:00
Pavel 79b5ef9c14 Bulk ingest
Added a method based on indexGPT folder ingester. Additional rst reader included.
2023-02-10 19:44:42 +04:00
Patrick Shriwise 64fb36b3de Adding location argument to ingest scripts 2023-02-07 10:20:07 -06:00
Alex 08215248d7 Inputs folder change 2023-02-07 10:52:03 +00:00
monkish54 c94866e9e9 Add cost estimate feature
Calculates number of tokens/user cost and requires user permission to proceed.

User permission bypass is built-in to allow for non-human users.
2023-02-06 18:32:06 -08:00
Pavel 1c734727a1 Ingest rst with sphinx
Transforms all rst files in provided folder to txt format first (utilising sphinx library). In my tests size of raw sample decreased 2-3 times.
2023-02-06 23:43:23 +04:00
Alex b71a9bf5ee init2 2023-02-03 12:45:29 +00:00