Adding cuda kernel (optimized for sm80) for block-wise 4b quantized float 16 GEMM. #18619
Azure Pipelines / Linux GPU CI Pipeline (Linux_Build Linux_Build)
succeeded
Feb 28, 2024 in 15m 51s
Linux_Build Linux_Build succeeded
Loading