Commit Graph

39 Commits

Author SHA1 Message Date
Matt
f352005dcf
Fix: add missing properties (#530) 2023-07-06 09:52:47 +01:00
Matt
e931d29017
feat: new classes (#499)
* feat: new classes

* chore: private llm refactor and enable streaming

* fix: validation errors in pytest

* fix: default brain behavior

* Update backend/llm/base.py

* chore: remove 🪦

* fix: remove 🪦
2023-07-04 16:56:54 +01:00
Stan Girard
dcaa70a947
feat(functions): improved by removing only get_context (#459) 2023-07-03 12:34:37 +02:00
Matt
b33fcc8bf7
feat: base class for llms (#462) 2023-07-03 11:16:36 +02:00
Stan Girard
b76262c4f0
Fix a bunch of issues (#442)
* fix(tokens): overflow

* fix(chat): fixed deletion error of the chat

#441

* fix(default-brain): autoset default brain when on upload page

* fix(promptengineering): changed prompt a bit

* feat(growthbook): removed

* docs(readme): updated
2023-07-01 17:22:28 +02:00
Matt
6f047f4a39
feat: streaming for standard brain picking (#385)
* feat: streaming for standard brain picking

* fix(bug): private llm

* wip: test

Co-authored-by: Mamadou DICKO <mamadoudicko@users.noreply.github.com>

* wip: almost good

Co-authored-by: Mamadou DICKO <mamadoudicko@users.noreply.github.com>

* feat: useFetch

* chore: remove 💀

* chore: fix linting

* fix: forward the request if not streaming

* feat: streaming for standard brain picking

* fix(bug): private llm

* wip: test

Co-authored-by: Mamadou DICKO <mamadoudicko@users.noreply.github.com>

* wip: almost good

Co-authored-by: Mamadou DICKO <mamadoudicko@users.noreply.github.com>

* feat: useFetch

* chore: remove 💀

* chore: fix linting

* fix: forward the request if not streaming

* fix: 💀 code

* fix: check_user_limit

* feat: brain_id to new chat stream

* fix: missing imports

* feat: message_id created on backend

Co-authored-by: Mamadou DICKO <mamadoudicko@users.noreply.github.com>

* chore: remove dead

* remove: cpython

* remove: dead

---------

Co-authored-by: Mamadou DICKO <mamadoudicko@users.noreply.github.com>
2023-06-30 10:10:59 +02:00
Zineb El Bachiri
ccdc5bb7a7
Feat/multiple brains files (#361) 2023-06-28 19:39:27 +02:00
Stan Girard
81a3d48fbc
fix(vectorstore): issues with userid (#380)
* fix(vectorstore): issues with userid

* perf(analytics): added tracking for file upload and chat (#376)

* fix: conditionnaly update functions list (#379)

---------

Co-authored-by: Mamadou DICKO <63923024+mamadoudicko@users.noreply.github.com>
2023-06-26 19:02:03 +02:00
Mamadou DICKO
e045a55ae7
fix: conditionnaly update functions list (#379) 2023-06-26 17:27:06 +02:00
Stan Girard
5fc837b250
feat(brainpicking): simplified (#371)
* feat(functions): simplified

* refactor(openai): changed to brainpicking

* feat(functions): made them inherit from brainpicking

* feat(privatebrainpicking): added new class

* feat(history&context): added

* Delete test_brainpicking.py

* Delete __init__.py
2023-06-26 10:34:03 +02:00
Mamadou DICKO
59fe7b089b
feat(chat): use openai function for answer (#354)
* feat(chat): use openai function for answer (backend)

* feat(chat): use openai function for answer (frontend)

* chore: refacto BrainPicking

* feat: update chat creation logic

* feat: simplify chat system logic

* feat: set default method to gpt-3.5-turbo-0613

* feat: use user own openai key

* feat(chat): slightly improve prompts

* feat: add global error interceptor

* feat: remove unused endpoints

* docs: update chat system doc

* chore(linter): add unused import remove config

* feat: improve dx

* feat: improve OpenAiFunctionBasedAnswerGenerator prompt
2023-06-22 17:50:06 +02:00
Matt
83fde0aeea
feat: private llm (#360)
* feat: private llm

* Update backend/vectorstore/supabase.py

* Update backend/vectorstore/supabase.py
2023-06-22 09:45:35 +01:00
Mamadou DICKO
e1a740472f
Feat: chat name edit (#343)
* feat(chat): add name update

* chore(linting): add flake8

* feat: add chat name edit
2023-06-20 09:54:23 +02:00
Stan Girard
8ed8a2c9ef chore: 🤖 docs
added docs
2023-06-20 08:56:17 +02:00
Stan Girard
99258790ad refactor(brainpicking): removed one function 2023-06-19 23:14:42 +02:00
Stan Girard
969e0b48a8 refactor(brains): into brainpicking 2023-06-19 22:55:42 +02:00
Stan Girard
b8ec6fd342 refactor(brainsettings): added moved files 2023-06-19 21:21:26 +02:00
Stan Girard
d42f14f431 feat(brainpicking): new class 2023-06-19 20:51:13 +02:00
Stan Girard
17aaf18d61 feat(vectorstore): own folder 2023-06-19 20:15:34 +02:00
Stan Girard
954acf1286 fix(llm): changed to stuff 2023-06-19 17:53:07 +02:00
Stan Girard
225280f2f5 fix(doc): retrieval to 8 because it takes a long time 2023-06-19 11:23:58 +02:00
Stan Girard
c0ff030100 feat(k): increased to 16 2023-06-17 01:32:03 +02:00
Stan Girard
65c44b3fb9 feat(chat): improved chain 2023-06-17 01:16:11 +02:00
Stan Girard
f833de47bc feat(openai): 16k 2023-06-16 17:52:32 +02:00
Stan Girard
82c74186a8 fix(memory): added memory back 2023-06-15 15:25:12 +02:00
Stan Girard
9af6a250e3 fix(model): broke something yesterday in QA 2023-06-14 22:23:05 +02:00
Cezar Vasconcelos
3b7390dd61
Feat/add new openai llm models (#334)
* feat(frontend): update OpenAi models and maxToken select's

* fix: update openai model to use the most recent

* fix: adjust max token for gpt-3 turbo

* fix: duplicating models

* fix: openai model type list
2023-06-14 21:30:52 +02:00
Arthur Busser
4546e8bc1d
Improve the prompt sent to the LLM (#318)
This is a more detailed prompt that explains what Quivr is and what it
does.
2023-06-13 16:24:07 +02:00
Stan Girard
de8daa04bf fix(prompt): less picky 2023-06-13 14:43:53 +02:00
Stan Girard
6708a7860a fix(prompt): added prompt qa to anthropic 2023-06-13 11:00:15 +02:00
Stan Girard
ab7dfbbaec fix(prompt): condense prompt 2023-06-13 10:59:24 +02:00
Stan Girard
9f0ba8bce7 fix(prompt): fixed qa prompt 2023-06-13 10:35:06 +02:00
xleven
5a0f8326df
feat(chat): basic source documents support (#289) 2023-06-10 10:43:44 +02:00
Matt
a64df04eab
feat: self supplied key (#286) 2023-06-09 18:49:47 +02:00
Stan Girard
a3ca7ecb37
Back/refacto files (#240)
* feat(docker): added docker for prod

* feat(refacto): moved to modules
2023-06-03 23:12:42 +02:00
Stan Girard
72c92b1a54
VertexAI Google Cloud Palm2 Support (#226)
* feat(bard): added

* docs(readme): update

* chore(print): removed
2023-06-01 16:01:27 +02:00
Stan Girard
e0de23e24d
Feat/single brain (#215)
* feat(llm): update

* feat(singlebrain): added new table with user-id

* feat(user): get user from email

* feat(user_id): added search

*  add user_id to most endpoints

* docs(readme): new script

---------

Co-authored-by: gozineb <zinebe@theodo.fr>
2023-05-31 13:51:23 +02:00
shaun
a52fcbe684 add missing column 2023-05-21 23:59:06 -07:00
shaun
c38265a5f5 add summarization backend 2023-05-21 23:39:55 -07:00