2 Commits

Author	SHA1	Message	Date
Po Yen Chen	24b12d04af	[CK_TILE] fmha fwd splitkv optimization for decode (seqlen_q=1) (#1789) ... * Update license year * Add initial code to override decode problem * Fix splitkv traits/args overriding error * Reshape and transpose lse for decode * Remove debug code * Prettify example code * Use better function name * Add kMergeNumHeadGroupsSeqLenQ flag Kernel user can use this switch to turn on/off optimization for some problem sizes * Add missing flag declarations * Default turn off kMergeNumHeadGroupsSeqLenQ in codegen * Group similar statements together * Remove assumption of seqlen_q=1 * Remove kMergeNumHeadGroupsSeqLenQ from splitkv combine kernel * Support kMergeNumHeadGroupsSeqLenQ=true in fmha splitkv kernel * Run kMergeNumHeadGroupsSeqLenQ=true kernels when need * Fix group mode block skip logics * Undo changes of normal fwd kernel * Update in GridSize() and using GridSize() for splitkv kernel (#1799) --------- Co-authored-by: Qianfeng <qianfeng.zhang@amd.com>	2025-01-07 18:49:24 +08:00
feli	4bc610416a	Ck tile/layernorm: implement naive reduce, opt performance (#1784) ... * add no welford * enable output raw * raw of int8 * fix build * fix smoke test err * [ck_tile]layernorm: fix welford ok, set int8 and bf16 small N as default and others open by generate * [cktile]layernorm, fix err commit files and remove uselss * fix quant 8192 err & change norm_reduce class and file name --------- Co-authored-by: coderfeli <coderfeli@163.com> Co-authored-by: carlushuang <carlus.huang@amd.com>	2025-01-03 14:28:59 +08:00