Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Added search quality testing pipeline #1774

Merged
merged 11 commits into from
Jul 6, 2024
Merged

Conversation

hagen-danswer
Copy link
Collaborator

No description provided.

Copy link

vercel bot commented Jul 3, 2024

The latest updates on your projects. Learn more about Vercel for Git ↗︎

Name Status Preview Comments Updated (UTC)
internal-search ✅ Ready (Inspect) Visit Preview 💬 Add feedback Jul 6, 2024 6:50pm

Copy link
Contributor

@yuhongsun96 yuhongsun96 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Generally looks good, didn't scrutinize the scripts that much but seems reasonable.

By the way, there are a ton of errors in the automated checks. I know this is more of a [WIP] PR but let's definitely fix those up before the final PR

backend/danswer/configs/app_configs.py Outdated Show resolved Hide resolved
backend/danswer/db/engine.py Outdated Show resolved Hide resolved
backend/danswer/document_index/vespa/index.py Outdated Show resolved Hide resolved
backend/tests/regression/answer_quality/relari.py Outdated Show resolved Hide resolved
@@ -0,0 +1,406 @@
version: '3'
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I'm quite against having yet another docker compose yaml. Every time we add a new env based config we have to currently copy it to 3 docker compose yamls, kubernetes deployment, helm deployment. Let's see if we can just use the existing yamls

ports:
- "8080"
environment:
# Auth Settings
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

this can rely on an .env file explicitly. The goal is to not have to duplicate the env vars so that when we introduce a new one it has to be changed in 20 places

o: bind
device: ${DANSWER_VESPA_DATA_DIR:-./vespa_data}
model_cache_huggingface:
# driver: local
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

general cleanup here (also other commented things around the PR). Be sure to give the PR a read on GitHub before the final review <3 thanks!

# driver_opts:
# type: none
# o: bind
# device: ${DANSWER_MODEL_CACHE_DIR:-./model_cache}
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

NEWLINE REEEE

- LANGUAGE_CHAT_NAMING_HINT=${LANGUAGE_CHAT_NAMING_HINT:-}
- QA_PROMPT_OVERRIDE=${QA_PROMPT_OVERRIDE:-}
# Other Services
- POSTGRES_HOST=relational_db
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Some of the env things do still need to stay as it points to a service in the docker compose deployment which is not default

Copy link
Contributor

@yuhongsun96 yuhongsun96 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Super amazing work! Thanks!

@yuhongsun96 yuhongsun96 merged commit ac14369 into main Jul 6, 2024
8 checks passed
@yuhongsun96 yuhongsun96 deleted the build-search-testing-pipeline branch July 6, 2024 18:51
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

None yet

2 participants