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

[LLM] Add code llama example #3050

Merged
merged 52 commits into from
Feb 1, 2024
Merged

[LLM] Add code llama example #3050

merged 52 commits into from
Feb 1, 2024

Conversation

Michaelvll
Copy link
Collaborator

@Michaelvll Michaelvll commented Jan 30, 2024

After #3048 is fixed, we should be able to make the model path a env var.

Tested (run the relevant ones):

  • Code formatting: bash format.sh
  • Any manual or new tests for this PR (please specify below)
    • Two commands in the readme
  • All smoke tests: pytest tests/test_smoke.py
  • Relevant individual smoke tests: pytest tests/test_smoke.py::test_fill_in_the_name
  • Backward compatibility tests: bash tests/backward_comaptibility_tests.sh

Copy link
Collaborator

@romilbhardwaj romilbhardwaj left a comment

Choose a reason for hiding this comment

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

Nice! Thanks @Michaelvll!! Left some comments. Tried out the sky launch bit, works nicely. Trying out with sky serve now.

llm/codellama/README.md Outdated Show resolved Hide resolved
llm/codellama/README.md Outdated Show resolved Hide resolved
llm/codellama/README.md Outdated Show resolved Hide resolved
llm/codellama/README.md Outdated Show resolved Hide resolved
llm/codellama/README.md Outdated Show resolved Hide resolved
llm/codellama/README.md Outdated Show resolved Hide resolved
llm/codellama/README.md Outdated Show resolved Hide resolved
llm/codellama/README.md Outdated Show resolved Hide resolved
llm/codellama/complete.py Outdated Show resolved Hide resolved
llm/codellama/README.md Show resolved Hide resolved
llm/codellama/README.md Outdated Show resolved Hide resolved
llm/codellama/README.md Outdated Show resolved Hide resolved
llm/codellama/README.md Outdated Show resolved Hide resolved
```
As shown, the service is now backed by 2 replicas, one on Azure and one on GCP, and the accelerator
type is chosen to be **the cheapest and available one** on the clouds. That said, it maximizes the
availability of the service while minimizing the cost.
Copy link
Collaborator

@MaoZiming MaoZiming Jan 31, 2024

Choose a reason for hiding this comment

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

What does it mean by 'maximize availability' here? Does it mean finding the instance faster?

Copy link
Collaborator Author

Choose a reason for hiding this comment

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

More like it increases the chance of getting the resources with all the candidate resources in multiple resource pools.

Michaelvll and others added 12 commits January 31, 2024 01:30
Co-authored-by: Romil Bhardwaj <romil.bhardwaj@berkeley.edu>
Co-authored-by: Romil Bhardwaj <romil.bhardwaj@berkeley.edu>
Co-authored-by: Ziming Mao <ziming.mao@yale.edu>
Co-authored-by: Ziming Mao <ziming.mao@yale.edu>
Co-authored-by: Ziming Mao <ziming.mao@yale.edu>
Co-authored-by: Ziming Mao <ziming.mao@yale.edu>
Copy link
Collaborator

@romilbhardwaj romilbhardwaj left a comment

Choose a reason for hiding this comment

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

LGTM, thanks @Michaelvll! This is very cool!

* No one else sees your chat history


![Code Llama GUI](https://imgur.com/Dor1MoE.gif "Chatting with Code Llama")
Copy link
Collaborator

Choose a reason for hiding this comment

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

Should we put this GIF at the top, say at L2 to catch eyes?

@@ -38,7 +38,6 @@ run: |

python3 -m fastchat.serve.controller --host 0.0.0.0 --port ${CONTROLLER_PORT} > ~/controller.log 2>&1 &

cd FastChat
Copy link
Collaborator

Choose a reason for hiding this comment

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

I am assuming this is intended : )

Copy link
Collaborator Author

Choose a reason for hiding this comment

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

Yep, it seems we forgot to remove this line before, when we change the dependency.

llm/codellama/README.md Outdated Show resolved Hide resolved
llm/codellama/README.md Outdated Show resolved Hide resolved
llm/codellama/README.md Outdated Show resolved Hide resolved
llm/codellama/gui.yaml Outdated Show resolved Hide resolved
llm/codellama/gui.yaml Show resolved Hide resolved
Co-authored-by: Romil Bhardwaj <romil.bhardwaj@berkeley.edu>
@Michaelvll
Copy link
Collaborator Author

Added the example for tabby. PTAL @romilbhardwaj @MaoZiming : )

@Michaelvll Michaelvll merged commit b25471c into master Feb 1, 2024
19 checks passed
@Michaelvll Michaelvll deleted the codellama branch February 1, 2024 16:21
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

4 participants