AlpinDale e14223dce5 kernel: use `cub::BlockReduce` instead of custom impl (#895) | 3 주 전 | |
---|---|---|
.. | ||
amd | f1d0b77c92 [0.6.0] Release Candidate (#481) | 4 달 전 |
nvidia | f1d0b77c92 [0.6.0] Release Candidate (#481) | 4 달 전 |
common.cu | e14223dce5 kernel: use `cub::BlockReduce` instead of custom impl (#895) | 3 주 전 |
fp8_marlin.cu | f1d0b77c92 [0.6.0] Release Candidate (#481) | 4 달 전 |