Skip to content

Adding cuda kernel (optimized for sm80) for block-wise 4b quantized float 16 GEMM. #30287

Adding cuda kernel (optimized for sm80) for block-wise 4b quantized float 16 GEMM.

Adding cuda kernel (optimized for sm80) for block-wise 4b quantized float 16 GEMM. #30287

Triggered via pull request January 30, 2024 17:42
Status Success
Total duration 23s
Artifacts
Validation
12s
Validation
Fit to window
Zoom out
Zoom in

Annotations

1 warning
Validation
Node.js 16 actions are deprecated. Please update the following actions to use Node.js 20: gradle/wrapper-validation-action@v1. For more information see: https://github.blog/changelog/2023-09-22-github-actions-transitioning-from-node-16-to-node-20/.