WrenAI

opensource/WrenAI

Fork 0

mirror of https://github.com/Canner/WrenAI.git synced 2026-09-24 23:29:49 +08:00

Files

T

History

9507838991 Feature(ai service): refine ask/ask details pipelines (#64 )

* allow visualization using one file

* remove default argparse option value

* resolve conflict

* resolve conflict

* resolve conflict

* update setup instructions

* resolve conflict

* resolve conflicts

* resolve conflict

* resolve conflict

* resolve conflict

* add anthropic model pricing

* allow change model for anthropic

* resolve conflict

* resolve conflict

* resovle conflict

* resovle conflict

* resolve conflict

* resolve conflict

* add container:ubuntu

* undo

* resolve conflict

* resolve conflict

* resolve conflict

* resolve conflict

* resolve conflict

* update

* resolve conflict

* resolve conflicts

* resolve conflicts

* resolve conflicts

* resolve conflicts

* resolve conflicts

* resolve conflicts

* resolve conflicts

* resolve conflicts

* remove unused files

* resolve conflict

* resolve conflict

* generate ddl from mdl for indexing

* resolve conflict

* resolve conflicts

* eval ask with semantic description (#45)

* update make eval-ask

* add semantic description to mdl before eval

* refine

* add model semantic generation

* simplify make command

* print timestamp

* minor update

---------

Co-authored-by: qa <qa@qadeMacBook-Pro.local>
Co-authored-by: ChihYu Yeh <chihyu.jimmy.yeh@gmail.com>
Co-authored-by: Aster Sun <aster.sun@cannerdata.com>

* remove ;

* simplify prompt

* revert topk

* update prompt

* make default top_k for retriever 10

* generate multiple results for ask pipeline

* resolve conflict

* add followup_generation_pipeline

* add logging for ask api results

* fix generate_mdl issue

* fix ask_details error message

* add extra standard package for uvicorn

* refine ask

* fix postprocessor for ask details

* fix preview data

* fix test

* update prompts

* refine prompts for ask and ask details

* fix prompt

* refine prompt

* remove slash at the end of api endpoints

* refine prompt

* add custom semantic run option

* resolve conflict

* fix bugs for tests and eval and restructure eval-ask

* fix trainling slash and add logging

* remove unused code

* restructure eval ask pipeline outputs

* add comments to the eval command

* change generation component argument name

* add try/except to handle other errors that might happen

* change WREN_AI_SERVICE_VERSION to nightly

* fix eval ask_details errors

* add error message to indexing

---------

Co-authored-by: imAsterSun <61279528+imAsterSun@users.noreply.github.com>
Co-authored-by: qa <qa@qadeMacBook-Pro.local>
Co-authored-by: Aster Sun <aster.sun@cannerdata.com>
Co-authored-by: Pao Sheng <paooap.oappao@gmail.com>

2024-04-08 15:44:15 +08:00

bootstrap

Feature: build wren-ui docker image & docker-compose to start all service we need (#27 )

2024-03-22 16:39:48 +08:00

.env.example

Feature(ai service): refine ask/ask details pipelines (#64 )

2024-04-08 15:44:15 +08:00

docker-compose-dev.yaml

Chore/docker compose update (#69 )

2024-04-03 15:57:09 +08:00

docker-compose.yaml

Chore/docker compose update (#69 )

2024-04-03 15:57:09 +08:00

README.md

Feature/wren launcher (#33 )

2024-03-25 17:42:20 +08:00

README.md

Service

wren-engine: the engine service. check out example here: wren-engine /example
wren-ai-service: the AI service. check out example here: wren-ai-service docker-compose example
qdrant: the vector store ai service is using.
wren-ui: the UI service.
bootstrap: put required files to volume for engine service.

Volume

Shared data using data volume.

Path structure as following:

/mdl
- *.json (will put sample.json during bootstrap)
accounts
config.properties

Network

Check out Network drivers overview to learn more about bridge network driver.

How to start

copy .env.example to .env.local and modify the OpenAI API key.
(optional) if your port 3000 is occupied, you can modify the HOST_PORT in .env.local.
run docker-compose --env-file .env.local up to start all services.