jakpiase
jakpiase
While optimizing PP-HumanMatting model I have found that the computational graph contains some weird and non-optimal pattern. Instead of using `pad2d` op, there is a combination of `unsqueeze2` + `pad3d`...
Add unit tests for grouped gemm two stage covering all existing instances. Also made a fix to allow skipping empty gemms.
## Feature Hi, I am working at PaddlePaddle(chinese DL framework). We and other DL frameworks would extremely benefit from integrated bfloat16 numpy datatype. I have seen that TF added its...
Added structural sparsity blockwise gemm
## Proposed changes Add support for build ck tile examples package ## Checklist Please put an `x` into the boxes that apply. You can also fill these out after creating...
## Proposed changes Add 2:4 structured sparsity support for ck tile fp16 gemm ## Checklist Please put an `x` into the boxes that apply. You can also fill these out...
## Proposed changes Remove scratch usage from universal gemm by moving the if kbatch related condition oustide of kernel and passing memory operation enum as a template parameter ## Checklist...
## Proposed changes [CK_TILE] Add universal gemm mem skip A/B LDS pipelines for tall and skinny gemms. ## Checklist Please put an `x` into the boxes that apply. You can...