Weves
a099f8e296
Rework header a bit + remove assumption of all personas having a prompt
2023-12-14 23:06:39 -08:00
Weves
16c8969028
Chat UI
2023-12-14 22:18:42 -08:00
Yuhong Sun
65fde8f1b3
Chat Backend ( #801 )
2023-12-14 22:14:37 -08:00
Yuhong Sun
229db47e5d
Update LLM Key Check Logic ( #825 )
2023-12-09 13:41:31 -08:00
Weves
2e3397feb0
Check for slack bot token changes every 60 seconds
2023-12-08 14:14:22 -08:00
Weves
d5658ce477
Persona enhancements
2023-12-07 14:29:37 -08:00
Weves
ddf3f99da4
Add support for global API prefix env variable
2023-12-07 12:42:17 -08:00
Weves
56785e6065
Add model choice to Persona
2023-12-07 00:20:42 -08:00
Yuhong Sun
9e9a578921
Option to speed up DanswerBot by turning off chain of thought ( #816 )
2023-12-05 00:43:45 -08:00
Weves
f7172612e1
Allow persona usage for Slack bots
2023-12-04 19:20:03 -08:00
Yuhong Sun
5aa2de7a40
Fix Weak Models Concurrency Issue ( #811 )
2023-12-04 15:40:10 -08:00
Yuhong Sun
e0b87d9d4e
Fix Weak Model Prompt ( #810 )
2023-12-04 15:02:08 -08:00
Yuhong Sun
651de071f7
Improve English rephrasing for multilingual use case ( #808 )
2023-12-03 14:34:12 -08:00
John Bergvall
5629ca7d96
Copy SearchQuery model with updated attribute due to Config.frozen=True ( #806 )
...
Fixes the following TypeError:
api_server_1 | File "/usr/local/lib/python3.11/site-packages/anyio/to_thread.py", line 33, in run_sync
api_server_1 | return await get_asynclib().run_sync_in_worker_thread(
api_server_1 | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
api_server_1 | File "/usr/local/lib/python3.11/site-packages/anyio/_backends/_asyncio.py", line 877, in run_sync_in_worker_thread
api_server_1 | return await future
api_server_1 | ^^^^^^^^^^^^
api_server_1 | File "/usr/local/lib/python3.11/site-packages/anyio/_backends/_asyncio.py", line 807, in run
api_server_1 | result = context.run(func, *args)
api_server_1 | ^^^^^^^^^^^^^^^^^^^^^^^^
api_server_1 | File "/usr/local/lib/python3.11/site-packages/starlette/concurrency.py", line 53, in _next
api_server_1 | return next(iterator)
api_server_1 | ^^^^^^^^^^^^^^
api_server_1 | File "/app/danswer/utils/timing.py", line 47, in wrapped_func
api_server_1 | value = next(gen)
api_server_1 | ^^^^^^^^^
api_server_1 | File "/app/danswer/direct_qa/answer_question.py", line 243, in answer_qa_query_stream
api_server_1 | top_chunks = cast(list[InferenceChunk], next(search_generator))
api_server_1 | ^^^^^^^^^^^^^^^^^^^^^^
api_server_1 | File "/app/danswer/search/search_runner.py", line 469, in full_chunk_search_generator
api_server_1 | retrieved_chunks = retrieve_chunks(
api_server_1 | ^^^^^^^^^^^^^^^^
api_server_1 | File "/app/danswer/search/search_runner.py", line 353, in retrieve_chunks
api_server_1 | q_copy.query = rephrase
api_server_1 | ^^^^^^^^^^^^
api_server_1 | File "pydantic/main.py", line 359, in pydantic.main.BaseModel.__setattr__
api_server_1 | TypeError: "SearchQuery" is immutable and does not support item assignment
2023-12-03 13:47:11 -08:00
Yuhong Sun
bc403d97f2
Organize Prompts for Chat implementation ( #807 )
2023-12-03 13:27:11 -08:00
Yuhong Sun
02095e9281
Restructure APIs ( #803 )
2023-12-02 14:48:08 -08:00
Yuhong Sun
8954a04602
Reorder Tables for cleaner extending ( #800 )
2023-12-01 17:46:13 -08:00
Yuhong Sun
8020db9e9a
Update connector interface with optional Owners information ( #798 )
2023-11-30 23:08:16 -08:00
Yuhong Sun
17c2f06338
Add more metadata options for File connector ( #797 )
2023-11-30 13:24:22 -08:00
Weves
9cff294a71
Increase retries for google drive connector
2023-11-30 03:03:26 -08:00
Weves
e983aaeca7
Add more logging on existing jobs
2023-11-30 02:58:37 -08:00
Yuhong Sun
fda89ac810
Expert Recommendation Heuristic Only ( #791 )
2023-11-29 15:53:57 -08:00
Yuhong Sun
006fd4c438
Ingestion API now always updates regardless of document updated_at ( #786 )
2023-11-29 02:08:50 -08:00
Weves
c64c25b2e1
Fix temp file deletion
2023-11-29 02:00:20 -08:00
Yuhong Sun
c2727a3f19
Custom OpenAI Model Server ( #782 )
2023-11-29 01:41:56 -08:00
Chris Weaver
37daf4f3e4
Remove AI Thoughts by default ( #783 )
...
- Removes AI Thoughts by default - only shows when validation fails
- Removes punctuation "words" from queries in addition to stopwords (Vespa ignores punctuation anyways)
- Fixes Vespa deletion script for larger doc counts
2023-11-29 01:00:53 -08:00
Yuhong Sun
fcb7f6fcc0
Accept files with character issues ( #781 )
2023-11-28 22:43:58 -08:00
Yuhong Sun
187b94a7d8
Blurb Key Error ( #778 )
2023-11-28 16:09:33 -08:00
Weves
30225fd4c5
Fix filter hiding
2023-11-28 04:13:11 -08:00
Weves
eab4fe83a0
Remove Slack bot personas from web UI
2023-11-28 02:53:18 -08:00
Chris Weaver
78d1ae0379
Customizable personas ( #772 )
...
Also includes a small fix to LLM filtering when combined with reranking
2023-11-28 00:57:48 -08:00
Yuhong Sun
87beb1f4d1
Log LLM details on server start ( #773 )
2023-11-27 21:32:48 -08:00
Yuhong Sun
05c2b7d34e
Update LLM related Libs ( #771 )
2023-11-26 19:54:16 -08:00
Yuhong Sun
39d09a162a
Danswer APIs Document Ingestion Endpoint ( #716 )
2023-11-26 19:09:22 -08:00
Yuhong Sun
d291fea020
Turn off Reranking for Streaming Flows ( #770 )
2023-11-26 16:45:23 -08:00
Yuhong Sun
2665bff78e
Option to turn off LLM for eval script ( #769 )
2023-11-26 15:31:03 -08:00
Yuhong Sun
65d38ac8c3
Slack to respect LLM chunk filter settings ( #768 )
2023-11-26 01:06:12 -08:00
Yuhong Sun
8391d89bea
Fix Indexing Concurrency ( #767 )
2023-11-25 21:40:36 -08:00
Yuhong Sun
ac2ed31726
Indexing Jobs to have shorter lived DB sessions ( #766 )
2023-11-24 21:38:16 -08:00
Chris Weaver
47f947b045
Use torch.multiprocessing + enable SimpleJobClient by default ( #765 )
2023-11-24 18:29:28 -08:00
Weves
3cec854c5c
Allow different model servers for different models / indexing jobs
2023-11-23 23:39:03 -08:00
Weves
26c6651a03
Improve LLM answer parsing
2023-11-23 15:03:35 -08:00
Yuhong Sun
13001ede98
Search Regression Test and Save/Load State updates ( #761 )
2023-11-23 00:00:30 -08:00
Yuhong Sun
fda377a2fa
Regression Script for Search quality ( #760 )
2023-11-22 19:33:28 -08:00
Yuhong Sun
bdfb894507
Slack Role Override ( #755 )
2023-11-22 17:47:18 -08:00
Weves
35c3511daa
Increase Vespa timeout
2023-11-22 01:42:59 -08:00
Chris Weaver
c1e19d0d93
Add selected docs in UI + rework the backend flow a bit( #754 )
...
Changes the flow so that the selected docs are sent over in a separate packet rather than as part of the initial packet for the streaming QA endpoint.
2023-11-21 19:46:12 -08:00
mattboret
e78aefb408
Add script to analyse the sources selection ( #721 )
...
---------
Co-authored-by: Matthieu Boret <matthieu.boret@fr.clara.net>
2023-11-21 18:35:26 -08:00
Bryan Peterson
aa2e859b46
add missing dependencies in model_server dockerfile ( #752 )
...
Thanks for catching this! Super helpful!
2023-11-21 17:59:28 -08:00
Yuhong Sun
c0c8ae6c08
Minor Tuning for Filters ( #753 )
2023-11-21 15:47:58 -08:00
