Skip to content

Adding cuda kernel (optimized for sm80) for block-wise 4b quantized float 16 GEMM. #30037

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. #30037

# Check gradle-wrapper.jar checksums against official Gradle releases.
# See: https://github.com/marketplace/actions/gradle-wrapper-validation
# This workflow was copied from the link above.
name: "Validate Gradle Wrapper"
on: [push, pull_request]
jobs:
validation:
name: "Validation"
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
- uses: gradle/wrapper-validation-action@v1