mirror of
https://github.com/ROCm/composable_kernel.git
synced 2026-05-04 21:51:28 +00:00
update layernorm (#1570)
* port layernorm
* change warp_welford.hpp
* Update warpshuffle
* 1. Add save mean and save std back
2. Move construction of tensor_view and tile_window to operator()
* refine welford max count calculation
* unify layernorm api
* Rename file
* Remove save mean and inv std
* Revert "refine welford max count calculation"
This reverts commit 022365802b.
* Fix order of parameter
* refine welford max count calculation again
* Remove fp32 instances
* Fix bug of padding
* refactor api
* Support bf16
* Extract common function
* Refine arg of operator()
* Add kMThreadPerBlock to template parameter
* clang format
* Refine variable name
* Refine file name
* remove redundant line
* refactor layernorm2d pipeline and add block-per-block utility
* fix name
* rename more
* add more block-per-tile instance
* remove duplicated define
* update instance for 2048, 1024 case
* support up to 2048 now
* opt loading
* add n1536
* Add two pass pipeline
* format
* Fix incorrect type
* parallel compilation
* Use smaller N
* fix 2p pass
* Support Repeat_M in distribution
* Refine nameing
* Add reduce example
---------
Co-authored-by: letaoqin <letaoqin@amd.com>
Co-authored-by: aska-0096 <haocwang@amd.com>
Co-authored-by: rocking <ChunYu.Lai@amd.com>
Co-authored-by: carlushuang <carlus.huang@amd.com>
This commit is contained in:
ltqin
362
include/ck_tile/ops/welford/block/block_welford.hpp
Normal file
362
include/ck_tile/ops/welford/block/block_welford.hpp
Normal file
@@ -0,0 +1,362 @@
|
||||
// SPDX-License-Identifier: MIT
|
||||
// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
|
||||
|
||||
#pragma once
|
||||
|
||||
#include "ck_tile/core.hpp"
|
||||
#include "ck_tile/ops/welford/thread/thread_welford.hpp"
|
||||
|
||||
namespace ck_tile {
|
||||
|
||||
template <typename Problem_, typename Policy_ = void>
|
||||
struct BlockWelford
|
||||
{
|
||||
using Problem = remove_cvref_t<Problem_>;
|
||||
using XDataType = typename Problem::XDataType;
|
||||
using ComputeDataType = typename Problem::ComputeDataType;
|
||||
|
||||
CK_TILE_DEVICE constexpr BlockWelford() {}
|
||||
|
||||
// [CAUSION] - max_count_ is to deal with the padding problem
|
||||
// max_count_ is depend on caller, eg: naive and splitN welford will have different
|
||||
// calculation of max_count_
|
||||
// -> use block_welford_calculate_max_count to compute
|
||||
template <typename XDistributedTensor_,
|
||||
typename MeanDistributedTensor_,
|
||||
typename VarDistributedTensor_>
|
||||
CK_TILE_DEVICE void operator()(const XDistributedTensor_& x_tensor,
|
||||
MeanDistributedTensor_& mean_tensor,
|
||||
VarDistributedTensor_& var_tensor,
|
||||
int& cur_count_, // -> prefer init as zero
|
||||
const int& max_count_)
|
||||
{
|
||||
constexpr auto I0 = number<0>{};
|
||||
constexpr auto I1 = number<1>{};
|
||||
|
||||
constexpr auto spans = XDistributedTensor_::get_distributed_spans();
|
||||
|
||||
sweep_tile_span(spans[I1], [&](auto dstr_idx_i1) {
|
||||
if(cur_count_ < max_count_)
|
||||
{
|
||||
++cur_count_;
|
||||
|
||||
sweep_tile_span(spans[I0], [&](auto dstr_idx_i0) {
|
||||
constexpr auto in_dstr_idx = make_tuple(dstr_idx_i0, dstr_idx_i1);
|
||||
constexpr auto out_dstr_idx = make_tuple(dstr_idx_i0);
|
||||
|
||||
auto x = ck_tile::type_convert<ComputeDataType>(x_tensor[in_dstr_idx]);
|
||||
|
||||
welford_update(
|
||||
mean_tensor(out_dstr_idx), var_tensor(out_dstr_idx), x, cur_count_);
|
||||
});
|
||||
}
|
||||
});
|
||||
}
|
||||
|
||||
template <typename XDistributedTensor_>
|
||||
CK_TILE_DEVICE static auto MakeMeanVarBlockTile()
|
||||
{
|
||||
static_assert(std::is_same_v<XDataType, typename XDistributedTensor_::DataType>, "wrong!");
|
||||
|
||||
constexpr auto reduce_dims = sequence<1>{};
|
||||
|
||||
constexpr auto dstr =
|
||||
make_static_tile_distribution(detail::make_reduce_tile_distribution_encoding(
|
||||
XDistributedTensor_::get_tile_distribution()
|
||||
.get_static_tile_distribution_encoding(),
|
||||
reduce_dims));
|
||||
|
||||
auto tensor = make_static_distributed_tensor<ComputeDataType>(dstr);
|
||||
|
||||
return tensor;
|
||||
}
|
||||
|
||||
template <typename XDistributedTensor_>
|
||||
CK_TILE_DEVICE auto
|
||||
operator()(const XDistributedTensor_& x_tensor, int& cur_count_, const int& max_count_)
|
||||
{
|
||||
auto mean_tensor = MakeMeanVarBlockTile<XDistributedTensor_>();
|
||||
auto var_tensor = MakeMeanVarBlockTile<XDistributedTensor_>();
|
||||
clear_tile(mean_tensor);
|
||||
clear_tile(var_tensor);
|
||||
|
||||
(*this)(x_tensor, mean_tensor, var_tensor, cur_count_, max_count_);
|
||||
|
||||
return ck_tile::make_tuple(mean_tensor, var_tensor);
|
||||
}
|
||||
};
|
||||
|
||||
template <typename Problem_, typename Policy_ = void>
|
||||
struct BlockWelfordSync
|
||||
{
|
||||
using Problem = remove_cvref_t<Problem_>;
|
||||
|
||||
template <typename MeanDistributedTensor_, typename VarDistributedTensor_>
|
||||
CK_TILE_DEVICE void
|
||||
operator()(MeanDistributedTensor_& mean_tensor, VarDistributedTensor_& var_tensor, int& count)
|
||||
{
|
||||
using Dstr = typename MeanDistributedTensor_::StaticTileDistribution;
|
||||
using DstrEncode = typename Dstr::DstrEncode;
|
||||
using DstrEncodeDetail = typename DstrEncode::detail;
|
||||
|
||||
static_assert(std::is_same_v<Dstr, typename VarDistributedTensor_::StaticTileDistribution>,
|
||||
"wrong!");
|
||||
|
||||
constexpr index_t NDimP = Dstr::get_num_of_dimension_p();
|
||||
constexpr index_t NDimR = Dstr::get_num_of_dimension_r();
|
||||
|
||||
constexpr index_t idim_p_lane = NDimP - 1;
|
||||
|
||||
// const auto ps_idx = make_array<index_t>(get_warp_id(), get_lane_id());
|
||||
// const auto rs_idx =
|
||||
// mean_tensor.get_tile_distribution().calculate_rs_index_from_ps_index(ps_idx);
|
||||
|
||||
constexpr index_t thread_buf_size = MeanDistributedTensor_::get_thread_buffer_size();
|
||||
static_assert(thread_buf_size == VarDistributedTensor_::get_thread_buffer_size());
|
||||
|
||||
const int original_count = count;
|
||||
|
||||
// loop over thread data
|
||||
static_for<0, thread_buf_size, 1>{}([&](auto i) {
|
||||
auto v_local_mean = mean_tensor.get_thread_buffer()[i];
|
||||
auto v_local_var = var_tensor.get_thread_buffer()[i];
|
||||
auto v_local_count = original_count;
|
||||
|
||||
// cross-lane reduce for replication
|
||||
// only reduce on R dimension correspond to lane
|
||||
// (lane id maps to this R dimension)
|
||||
static_for<0, NDimR, 1>{}([&](auto idim_r) {
|
||||
// FIXME: nasty to use does_p_own_r_
|
||||
if constexpr(DstrEncodeDetail::does_p_own_r_[idim_p_lane][idim_r])
|
||||
{
|
||||
constexpr index_t r_length = DstrEncode::rs_lengths_[idim_r];
|
||||
|
||||
constexpr index_t lid_over_rid_derivative =
|
||||
DstrEncodeDetail::ps_over_rs_derivative_[idim_p_lane][idim_r];
|
||||
|
||||
static_assert(is_power_of_two_integer(r_length),
|
||||
"wrong! only support power of 2 reduction");
|
||||
|
||||
constexpr index_t nstage = integer_log2_floor(r_length);
|
||||
|
||||
// reduction sweep forward
|
||||
static_for<0, nstage, 1>{}([&](auto istage) {
|
||||
// xor
|
||||
index_t src_lane =
|
||||
(__lane_id()) ^
|
||||
(number<lid_over_rid_derivative << istage.value>{}.value);
|
||||
|
||||
// pull data from remote lane
|
||||
const auto v_remote_mean = warp_shuffle(v_local_mean, src_lane);
|
||||
const auto v_remote_var = warp_shuffle(v_local_var, src_lane);
|
||||
const auto v_remote_count = warp_shuffle(v_local_count, src_lane);
|
||||
|
||||
// welford merge
|
||||
welford_merge(v_local_mean,
|
||||
v_local_var,
|
||||
v_local_count,
|
||||
v_remote_mean,
|
||||
v_remote_var,
|
||||
v_remote_count);
|
||||
});
|
||||
}
|
||||
});
|
||||
|
||||
mean_tensor.get_thread_buffer()(i) = v_local_mean;
|
||||
var_tensor.get_thread_buffer()(i) = v_local_var;
|
||||
|
||||
count = v_local_count;
|
||||
});
|
||||
}
|
||||
};
|
||||
|
||||
template <typename Problem_, typename Policy_ = void>
|
||||
struct BlockWelfordCrossWarpSync
|
||||
{
|
||||
using Problem = remove_cvref_t<Problem_>;
|
||||
using BlockShape = typename Problem::BlockShape;
|
||||
|
||||
template <typename MeanDistributedTensor_>
|
||||
CK_TILE_DEVICE static constexpr index_t GetReduceWarps()
|
||||
{
|
||||
constexpr index_t num_reduce_warps = [&]() {
|
||||
using Dstr = typename MeanDistributedTensor_::StaticTileDistribution;
|
||||
using DstrEncode = typename Dstr::DstrEncode;
|
||||
using DstrEncodeDetail = typename DstrEncode::detail;
|
||||
|
||||
constexpr index_t NDimR = Dstr::get_num_of_dimension_r();
|
||||
|
||||
constexpr index_t idim_p_warp = 0;
|
||||
|
||||
index_t len_ = 1;
|
||||
static_for<0, NDimR, 1>{}([&](auto idim_r) {
|
||||
if constexpr(DstrEncodeDetail::does_p_own_r_[idim_p_warp][idim_r])
|
||||
{
|
||||
constexpr index_t r_length = DstrEncode::rs_lengths_[idim_r];
|
||||
len_ *= r_length;
|
||||
}
|
||||
});
|
||||
return len_;
|
||||
}();
|
||||
return num_reduce_warps;
|
||||
}
|
||||
|
||||
// return in byte
|
||||
template <typename MeanDistributedTensor_>
|
||||
CK_TILE_HOST_DEVICE static constexpr index_t GetSmemSize()
|
||||
{
|
||||
// constexpr auto num_reduce_warps = GetReduceWarps<MeanDistributedTensor_>();
|
||||
|
||||
// data need to exchange is very small, we just pack mean+var+count -> 4dword
|
||||
constexpr index_t thread_buf_size = MeanDistributedTensor_::get_thread_buffer_size();
|
||||
|
||||
// we need to store all data from every wave into smem
|
||||
// e.g. 2x2 reduce along N
|
||||
// -------------> reduce N
|
||||
// | w0 | w1 | ___> | w01 |
|
||||
// | w2 | w3 | | w23 |
|
||||
//
|
||||
// -> store data from every wave into LDS
|
||||
//
|
||||
//
|
||||
// -------------> reduce N
|
||||
// | w0 | w1 | w2 | w3 | -----> | w0123 |
|
||||
//
|
||||
// -> also store data from every wave into LDS
|
||||
constexpr index_t num_warps = BlockShape::BlockSize / warpSize;
|
||||
return num_warps * 4 * thread_buf_size * sizeof(float);
|
||||
}
|
||||
|
||||
template <typename MeanDistributedTensor_, typename VarDistributedTensor_>
|
||||
CK_TILE_DEVICE void operator()(MeanDistributedTensor_& mean_tensor,
|
||||
VarDistributedTensor_& var_tensor,
|
||||
int& count,
|
||||
void* smem)
|
||||
{
|
||||
using DataType = typename MeanDistributedTensor_::DataType;
|
||||
using Dstr = typename MeanDistributedTensor_::StaticTileDistribution;
|
||||
// using DstrEncode = typename Dstr::DstrEncode;
|
||||
// using DstrEncodeDetail = typename DstrEncode::detail;
|
||||
|
||||
static_assert(std::is_same_v<Dstr, typename VarDistributedTensor_::StaticTileDistribution>,
|
||||
"wrong!");
|
||||
|
||||
constexpr index_t thread_buf_size = MeanDistributedTensor_::get_thread_buffer_size();
|
||||
static_assert(thread_buf_size == VarDistributedTensor_::get_thread_buffer_size());
|
||||
|
||||
// Note: we always pack everything into fp32x4
|
||||
fp32x4_t* smem_ptr = reinterpret_cast<fp32x4_t*>(smem);
|
||||
const index_t lane_id = get_lane_id();
|
||||
const index_t warp_id = get_warp_id();
|
||||
constexpr auto num_reduce_warps = GetReduceWarps<MeanDistributedTensor_>();
|
||||
constexpr index_t num_warps = BlockShape::BlockSize / warpSize;
|
||||
const index_t smem_offset = warp_id;
|
||||
|
||||
// skip if nonthing to do
|
||||
if constexpr(num_reduce_warps == 1)
|
||||
return;
|
||||
|
||||
// store into smem only for lane-0 within one warp
|
||||
if(lane_id == 0)
|
||||
{
|
||||
static_for<0, thread_buf_size, 1>{}([&](auto i) {
|
||||
fp32x4_t local_scratch_;
|
||||
local_scratch_[0] = bit_cast<float>(mean_tensor.get_thread_buffer()[i]);
|
||||
local_scratch_[1] = bit_cast<float>(var_tensor.get_thread_buffer()[i]);
|
||||
local_scratch_[2] = bit_cast<float>(count);
|
||||
|
||||
smem_ptr[smem_offset + i * num_warps] = local_scratch_;
|
||||
});
|
||||
}
|
||||
block_sync_lds();
|
||||
|
||||
// load from smem. here we let everythread to do compute :)
|
||||
index_t local_warp_id = warp_id / num_reduce_warps;
|
||||
index_t local_smem_os = local_warp_id * num_reduce_warps;
|
||||
fp32x4_t all_scratch[thread_buf_size * num_reduce_warps];
|
||||
static_for<0, thread_buf_size, 1>{}([&](auto i_0) {
|
||||
static_for<0, num_reduce_warps, 1>{}([&](auto i_1) {
|
||||
all_scratch[i_0 * num_warps + i_1] =
|
||||
smem_ptr[i_0 * num_reduce_warps + local_smem_os + i_1];
|
||||
});
|
||||
});
|
||||
block_sync_lds(); // TODO: we don't need sync here
|
||||
|
||||
// const int original_count = count;
|
||||
|
||||
static_for<0, thread_buf_size, 1>{}([&](auto i_0) {
|
||||
// TODO: use descriptor for this
|
||||
auto v_local = all_scratch[i_0 * num_warps];
|
||||
auto v_local_mean = bit_cast<DataType>(v_local[0]);
|
||||
auto v_local_var = bit_cast<DataType>(v_local[1]);
|
||||
auto v_local_count = bit_cast<int>(v_local[2]);
|
||||
|
||||
// further reduce mean/var
|
||||
static_for<0, num_reduce_warps - 1, 1>{}([&](auto i_1_n1) {
|
||||
constexpr auto i_1 = number<i_1_n1 + 1>{};
|
||||
const fp32x4_t v_remote = all_scratch[i_0 * num_warps + i_1];
|
||||
const auto v_remote_mean = bit_cast<DataType>(v_remote[0]);
|
||||
const auto v_remote_var = bit_cast<DataType>(v_remote[1]);
|
||||
const auto v_remote_count = bit_cast<int>(v_remote[2]);
|
||||
|
||||
welford_merge(v_local_mean,
|
||||
v_local_var,
|
||||
v_local_count,
|
||||
v_remote_mean,
|
||||
v_remote_var,
|
||||
v_remote_count);
|
||||
});
|
||||
|
||||
mean_tensor.get_thread_buffer()(i_0) = v_local_mean;
|
||||
var_tensor.get_thread_buffer()(i_0) = v_local_var;
|
||||
|
||||
count = v_local_count;
|
||||
});
|
||||
}
|
||||
};
|
||||
|
||||
// compute the max count for a last dim reduce
|
||||
// everything may have vector/repeat, so the max count could be uneven
|
||||
// TODO: specify which dim to compute and proper set the problem
|
||||
// TODO: BlockShape we reuse layernorm_fwd_shape :)
|
||||
template <typename BlockShape>
|
||||
CK_TILE_DEVICE constexpr index_t block_tile_welford_calculate_max_count(int row_size)
|
||||
{
|
||||
#if 0
|
||||
using S = BlockShape;
|
||||
index_t LastloopN = row_size % S::Block_N == 0 ? S::Block_N : row_size % S::Block_N;
|
||||
constexpr index_t NThread = S::WarpPerBlock_N * S::ThreadPerWarp_N;
|
||||
index_t iNLane = get_thread_id() % NThread;
|
||||
index_t iN0 = LastloopN / (S::Vector_N * S::ThreadPerWarp_N);
|
||||
index_t iN1 = (LastloopN % (S::Vector_N * S::ThreadPerWarp_N)) / S::Vector_N;
|
||||
index_t N2 = (LastloopN % (S::Vector_N * S::ThreadPerWarp_N)) % S::Vector_N;
|
||||
index_t iN3 = iNLane < iN1 ? S::Vector_N : iNLane == iN1 ? N2 : 0;
|
||||
return iN0 * S::Vector_N + iN3;
|
||||
#endif
|
||||
using S_ = BlockShape;
|
||||
constexpr index_t ThreadsPerBlock_N = S_::WarpPerBlock_N * S_::ThreadPerWarp_N;
|
||||
|
||||
// TODO: we always check vector size, need be evenly devidable by vector-n
|
||||
const index_t element_per_row = row_size / S_::Vector_N;
|
||||
index_t lane_id_n = get_thread_id() % ThreadsPerBlock_N;
|
||||
|
||||
index_t cnt = 0;
|
||||
// TODO: Repeat_N can not be too long, otherwise this is not good
|
||||
static_for<0, S_::Repeat_N, 1>{}([&](auto) {
|
||||
index_t _a = lane_id_n < element_per_row ? 1 : 0;
|
||||
cnt += _a;
|
||||
lane_id_n += ThreadsPerBlock_N;
|
||||
});
|
||||
return cnt * S_::Vector_N;
|
||||
}
|
||||
|
||||
// Note: this function must be called after all the computation
|
||||
template <typename VarDistributedTensor_>
|
||||
CK_TILE_DEVICE constexpr void block_tile_welford_post_scale_var(VarDistributedTensor_& var_tensor,
|
||||
int count)
|
||||
{
|
||||
using DataType = typename VarDistributedTensor_::DataType;
|
||||
tile_elementwise_inout([&count](auto& x) { x = x / type_convert<DataType>(count); },
|
||||
var_tensor);
|
||||
}
|
||||
} // namespace ck_tile
|
||||
18
include/ck_tile/ops/welford/block/block_welford_problem.hpp
Normal file
18
include/ck_tile/ops/welford/block/block_welford_problem.hpp
Normal file
@@ -0,0 +1,18 @@
|
||||
// SPDX-License-Identifier: MIT
|
||||
// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
|
||||
|
||||
#pragma once
|
||||
|
||||
#include "ck_tile/core.hpp"
|
||||
|
||||
namespace ck_tile {
|
||||
|
||||
template <typename XDataType_, typename ComputeDataType_, typename BlockShape_>
|
||||
struct BlockWelfordProblem
|
||||
{
|
||||
using XDataType = remove_cvref_t<XDataType_>;
|
||||
using ComputeDataType = remove_cvref_t<ComputeDataType_>;
|
||||
using BlockShape = remove_cvref_t<BlockShape_>;
|
||||
};
|
||||
|
||||
} // namespace ck_tile
|
||||
Reference in New Issue
Block a user