Adding cuda kernel (optimized for sm80) for block-wise 4b quantized float 16 GEMM. #18619
Azure Pipelines / Windows x64 QNN CI Pipeline
succeeded
Feb 29, 2024 in 44m 5s
Build #20240228.27 succeeded
Loading