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

chore: add NGL settings for vision models #4214

Merged
merged 1 commit into from
Dec 4, 2024

Conversation

louis-menlo
Copy link
Contributor

@louis-menlo louis-menlo commented Dec 4, 2024

Describe Your Changes

It’s an update to add new NGL settings for the llava model.json files. It’s a minor update for predefined models, but there should be a correct update from cortex.cpp to import vision models.

Related Issues

- #1763

Changes made

The changes in the code are as follows:

  1. package.json:

    • The version of the package @janhq/inference-cortex-extension has been updated from "1.0.23" to "1.0.24".
  2. model.json files:

    • Changes were made to three different model JSON files located in different subdirectories: bakllava-1, llava-13b, and llava-7b.
    • For each of these models, an additional property "ngl": 33 was added to the "model" key.
    • This update is consistent across all model files, suggesting a related change or parameter update applicable to multiple model configurations.

@github-actions github-actions bot added the type: chore Maintenance, operational label Dec 4, 2024
Copy link
Contributor

github-actions bot commented Dec 4, 2024

Barecheck - Code coverage report

Total: 69.29%

Your code coverage diff: 0.00% ▴

✅ All code changes are covered

@louis-menlo louis-menlo merged commit bf5d121 into dev Dec 4, 2024
11 checks passed
@louis-menlo louis-menlo deleted the chore/add-llava-ngl-in-model-json-files branch December 4, 2024 08:28
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
type: chore Maintenance, operational
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants