Adding cuda kernel (optimized for sm80) for block-wise 4b quantized float 16 GEMM. #18619
Azure Pipelines / Linux CPU ATen Pipeline
succeeded
Feb 29, 2024 in 11m 12s
Build #20240228.25 succeeded
Details
- Failed: 0 (0.00%)
- Passed: 8,896 (99.96%)
- Other: 4 (0.04%)
- Total: 8,900
Loading