squall/flash-attention
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Move dropout to a separate file (dropout.h)

Browse Source
This commit is contained in:
Tri Dao 2024-01-14 12:19:17 -08:00
parent 10dad61277
commit 1274ec3e7e
4 changed files with 112 additions and 95 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

91
csrc/flash_attn/src/dropout.h Normal file
View File

@ -0,0 +1,91 @@
#pragma once
#include "philox.cuh"
#include "utils.h"
namespace flash {
struct Dropout {
const unsigned long long seed, offset;
const uint8_t p_dropout_in_uint8_t;
inline __device__ Dropout(const unsigned long long seed, const unsigned long long offset,
const uint8_t p_dropout_in_uint8_t,
const int bid, const int hid, const int tid, const int nheads)
: seed(seed)
, offset(offset + (bid * nheads + hid) * 32 + tid % 32)
, p_dropout_in_uint8_t(p_dropout_in_uint8_t) {
}
template <bool encode_dropout_in_sign_bit=false, typename Engine, typename Layout>
inline __device__ void apply_dropout(Tensor<Engine, Layout> &tensor_,
int block_row_start, int block_col_start, int block_row_stride) {
// tensor_ has shape (nrow=(2, MMA_M), ncol=(2, MMA_N))
Tensor tensor = make_tensor(tensor_.data(), flash::convert_layout_rowcol_dropout(tensor_.layout()));
// tensor has shape (8, MMA_M, MMA_N / 2)
using T = typename Engine::value_type;
auto encode_dropout = [](bool keep, T val) {
return keep ? val : (encode_dropout_in_sign_bit ? -val : T(0));
};
static_assert(decltype(size<2>(tensor))::value % 2 == 0);
const uint16_t p_dropout_8bit_in_uint16_t = uint16_t(p_dropout_in_uint8_t);
const uint32_t p_dropout_8bit_in_uint32_t = (uint32_t(p_dropout_8bit_in_uint16_t) << 16) | uint32_t(p_dropout_8bit_in_uint16_t);
// if (cute::thread0()) { printf("threshold2 = 0x%x\n", p_dropout_8bit_in_uint32_t); }
#pragma unroll
for (int m = 0; m < size<1>(tensor); ++m, block_row_start += block_row_stride) {
uint2 rowcol = make_uint2(block_row_start, block_col_start);
#pragma unroll
for (int n = 0; n < size<2>(tensor) / 2; ++n, ++rowcol.y) {
// if (cute::thread(32, 0)) { printf("m = %d, n = %d, row = %d, col = %d\n", m, n, int(rowcol.x), int(rowcol.y));}
uint4 random_uint4 = flash::philox(seed, reinterpret_cast<unsigned long long&>(rowcol), offset);
// if (cute::thread0()) { printf("philox = %u, %d, %d, %d\n", random_uint4.x, random_uint4.y, random_uint4.z, random_uint4.w);}
uint8_t (&rnd_8)[16] = reinterpret_cast<uint8_t (&)[16]>(random_uint4);
// Special implementation for 16-bit types: we duplicate the threshold to the
// low and high 16 bits of a 32-bit value, then use the f16x2 comparison instruction
// to get a mask. The low 16 bits of the mask will be either 0xffff or 0x0000,
// and the high 16 bits will be either 0xffff or 0x0000, depending on whether
// the random value is less than the threshold.
// We then do a bit-wise AND between the mask and the original value (in 32-bit).
// We're exploiting the fact that floating point comparison is equivalent to integer
// comparison, since we're comparing unsigned integers whose top 8-bits are zero.
if (!encode_dropout_in_sign_bit
&& (std::is_same<T, cutlass::half_t>::value || std::is_same<T, cutlass::bfloat16_t>::value)) {
uint16_t rnd_16[16];
#pragma unroll
for (int i = 0; i < 16; i++) { rnd_16[i] = uint16_t(rnd_8[i]); }
uint32_t (&rnd_32)[8] = reinterpret_cast<uint32_t (&)[8]>(rnd_16);
#pragma unroll
for (int j = 0; j < 2; j++) {
Tensor tensor_uint32 = recast<uint32_t>(tensor(_, m, n * 2 + j));
// if (cute::thread0()) { printf("random = 0x%x, 0x%x, 0x%x, 0x%x\n", rnd_32[j * 4 + 0], rnd_32[j * 4 + 1], rnd_32[j * 4 + 2], rnd_32[j * 4 + 3]); }
// if (cute::thread0()) { printf("tensor_uint32 = 0x%x, 0x%x, 0x%x, 0x%x\n", tensor_uint32(0), tensor_uint32(1), tensor_uint32(2), tensor_uint32(3)); }
#pragma unroll
for (int i = 0; i < 4; i++) {
uint32_t mask;
asm volatile("set.le.u32.f16x2 %0, %1, %2;\n" : "=r"(mask) : "r"(rnd_32[j * 4 + i]), "r"(p_dropout_8bit_in_uint32_t));
tensor_uint32(i) &= mask;
}
// if (cute::thread0()) { printf("tensor_uint32 = 0x%x, 0x%x, 0x%x, 0x%x\n", tensor_uint32(0), tensor_uint32(1), tensor_uint32(2), tensor_uint32(3)); }
}
} else {
#pragma unroll
for (int j = 0; j < 2; j++) {
#pragma unroll
for (int i = 0; i < 8; i++) {
tensor(i, m, n * 2 + j) = encode_dropout(rnd_8[j * 8 + i] <= p_dropout_in_uint8_t, tensor(i, m, n * 2 + j));
}
Tensor tensor_uint32 = recast<uint32_t>(tensor(_, m, n * 2 + j));
// if (cute::thread0()) { printf("tensor_uint32 = 0x%x, 0x%x, 0x%x, 0x%x\n", tensor_uint32(0), tensor_uint32(1), tensor_uint32(2), tensor_uint32(3)); }
}
}
// // if ((threadIdx.x == 0) && (blockIdx.x == 0) && (blockIdx.y == 0)) {
// // printf("n = %d, ph Philox: %u, %u, %u, %u\n", n, rnd_8.x, rnd_8.y, rnd_8.z, rnd_8.w);
// // }
}
}
}
};
} // namespace flash

19
csrc/flash_attn/src/flash_bwd_kernel.h
View File

@ -14,6 +14,7 @@
#include "kernel_traits.h"
#include "utils.h"
#include "softmax.h"
#include "dropout.h"
#include "alibi.h"
@ -796,8 +797,8 @@ inline __device__ void compute_dq_dk_dv_1colblock(const Params &params, const in
cute::copy(smem_tiled_copy_KV, tdPsV, tdPrV_copy_view);
}
auto seed = params.rng_state[0];
auto offset = params.rng_state[1] + (bidb * params.h + bidh) * 32 + tidx % 32;
flash::Dropout dropout(params.rng_state[0], params.rng_state[1], params.p_dropout_in_uint8_t,
bidb, bidh, tidx, params.h);
clear(acc_dv);
clear(acc_dk);
@ -886,9 +887,8 @@ inline __device__ void compute_dq_dk_dv_1colblock(const Params &params, const in
// Need col to be multiples of 32, since we're doing dropout with block of 16 x 32
static_assert(MMA_N_SdP % 2 == 0);
int block_col_idx = n_block * (kBlockN / 32) + (warp_id / AtomLayoutMS) * (MMA_N_SdP / 2);
flash::apply_dropout</*encode_dropout_in_sign_bit=*/true>(
scores, params.p_dropout_in_uint8_t, seed, offset,
block_row_idx, block_col_idx, AtomLayoutMS
dropout.template apply_dropout</*encode_dropout_in_sign_bit=*/true>(
scores, block_row_idx, block_col_idx, AtomLayoutMS
);
}
// Convert scores from fp32 to fp16/bf16
@ -1395,8 +1395,8 @@ inline __device__ void compute_dq_dk_dv_1rowblock(const Params &params, const in
#pragma unroll
for (int mi = 0; mi < size(dP_sum); ++mi) { dP_sum(mi) = sdPsum(get<0>(taccScS_row(mi))); }
auto seed = params.rng_state[0];
auto offset = params.rng_state[1] + (bidb * params.h + bidh) * 32 + tidx % 32;
flash::Dropout dropout(params.rng_state[0], params.rng_state[1], params.p_dropout_in_uint8_t,
bidb, bidh, tidx, params.h);
clear(acc_dq);
@ -1445,9 +1445,8 @@ inline __device__ void compute_dq_dk_dv_1rowblock(const Params &params, const in
// Need col to be multiples of 32, since we're doing dropout with block of 16 x 32
static_assert(MMA_N_SdP % 2 == 0);
int block_col_idx = n_block * (kBlockN / 32) + (warp_id / AtomLayoutMS) * (MMA_N_SdP / 2);
flash::apply_dropout</*encode_dropout_in_sign_bit=*/true>(
scores, params.p_dropout_in_uint8_t, seed, offset,
block_row_idx, block_col_idx, AtomLayoutMS
dropout.template apply_dropout</*encode_dropout_in_sign_bit=*/true>(
scores, block_row_idx, block_col_idx, AtomLayoutMS
);
}
// Convert scores from fp32 to fp16/bf16

27
csrc/flash_attn/src/flash_fwd_kernel.h
View File

@ -14,6 +14,7 @@
#include "kernel_traits.h"
#include "utils.h"
#include "softmax.h"
#include "dropout.h"
#include "alibi.h"
@ -75,15 +76,15 @@ inline __device__ void compute_attn_1rowblock(const Params &params, const int bi
constexpr int kNWarps = Kernel_traits::kNWarps;
constexpr int MMA_M = kBlockM / decltype(size<0>(typename Kernel_traits::TiledMma::TiledShape_MNK{}))::value;
auto seeds = at::cuda::philox::unpack(params.philox_args);
unsigned long long seed = std::get<0>(seeds);
unsigned long long offset = std::get<1>(seeds) + (bidb * params.h + bidh) * 32 + tidx % 32;
auto seed_offset = at::cuda::philox::unpack(params.philox_args);
flash::Dropout dropout(std::get<0>(seed_offset), std::get<1>(seed_offset), params.p_dropout_in_uint8_t,
bidb, bidh, tidx, params.h);
// Save seed and offset for backward, before any early exiting. Otherwise the 0-th thread block might
// exit early and no one saves the rng states.
if (Is_dropout && blockIdx.x == 0 && blockIdx.y == 0 && blockIdx.z == 0 && tidx == 0) {
params.rng_state[0] = seed;
params.rng_state[1] = std::get<1>(seeds);
params.rng_state[0] = std::get<0>(seed_offset);
params.rng_state[1] = std::get<1>(seed_offset);
}
const BlockInfo</*Varlen=*/!Is_even_MN> binfo(params, bidb);
@ -404,16 +405,14 @@ inline __device__ void compute_attn_1rowblock(const Params &params, const int bi
if (Return_softmax) {
Tensor acc_s_f16 = flash::convert_type<Element>(acc_s);
Tensor acc_s_f16_drop = make_tensor(acc_s_f16.data(), rP.layout());
flash::apply_dropout</*encode_dropout_in_sign_bit=*/true>(
acc_s_f16_drop, params.p_dropout_in_uint8_t, seed, offset,
block_row_idx, block_col_idx, kNWarps
dropout.template apply_dropout</*encode_dropout_in_sign_bit=*/true>(
acc_s_f16_drop, block_row_idx, block_col_idx, kNWarps
);
cute::copy(acc_s_f16, tSgS);
tSgS.data() = tSgS.data() + (-kBlockN);
}
if (Is_dropout) {
flash::apply_dropout(rP, params.p_dropout_in_uint8_t, seed, offset,
block_row_idx, block_col_idx, kNWarps);
dropout.apply_dropout(rP, block_row_idx, block_col_idx, kNWarps);
}
// Reshape rP from (nrow=(2, MMA_M), ncol=(2, MMA_N)) to ((2, 2, 2), MMA_M, MMA_N / 2)
@ -489,16 +488,14 @@ inline __device__ void compute_attn_1rowblock(const Params &params, const int bi
if (Return_softmax) {
Tensor acc_s_f16 = flash::convert_type<Element>(acc_s);
Tensor acc_s_f16_drop = make_tensor(acc_s_f16.data(), rP.layout());
flash::apply_dropout</*encode_dropout_in_sign_bit=*/true>(
acc_s_f16_drop, params.p_dropout_in_uint8_t, seed, offset,
block_row_idx, block_col_idx, kNWarps
dropout.template apply_dropout</*encode_dropout_in_sign_bit=*/true>(
acc_s_f16_drop, block_row_idx, block_col_idx, kNWarps
);
cute::copy(acc_s_f16, tSgS);
tSgS.data() = tSgS.data() + (-kBlockN);
}
if (Is_dropout) {
flash::apply_dropout(rP, params.p_dropout_in_uint8_t, seed, offset,
block_row_idx, block_col_idx, kNWarps);
dropout.apply_dropout(rP, block_row_idx, block_col_idx, kNWarps);
}
// Reshape rP from (nrow=(2, MMA_M), ncol=(2, MMA_N)) to ((2, 2, 2), MMA_M, MMA_N / 2)

70
csrc/flash_attn/src/softmax.h
View File

@ -212,74 +212,4 @@ inline __device__ void apply_mask_causal_w_idx(
}
}
template <bool encode_dropout_in_sign_bit=false, typename Engine, typename Layout>
inline __device__ void apply_dropout(Tensor<Engine, Layout> &tensor_, uint8_t p_dropout_in_uint8_t,
unsigned long long seed, unsigned long long offset,
int block_row_start, int block_col_start,
int block_row_stride) {
// tensor_ has shape (nrow=(2, MMA_M), ncol=(2, MMA_N))
Tensor tensor = make_tensor(tensor_.data(), flash::convert_layout_rowcol_dropout(tensor_.layout()));
// tensor has shape (8, MMA_M, MMA_N / 2)
using T = typename Engine::value_type;
auto encode_dropout = [](bool keep, T val) {
return keep ? val : (encode_dropout_in_sign_bit ? -val : T(0));
};
static_assert(decltype(size<2>(tensor))::value % 2 == 0);
const uint16_t p_dropout_8bit_in_uint16_t = uint16_t(p_dropout_in_uint8_t);
const uint32_t p_dropout_8bit_in_uint32_t = (uint32_t(p_dropout_8bit_in_uint16_t) << 16) | uint32_t(p_dropout_8bit_in_uint16_t);
// if (cute::thread0()) { printf("threshold2 = 0x%x\n", p_dropout_8bit_in_uint32_t); }
#pragma unroll
for (int m = 0; m < size<1>(tensor); ++m, block_row_start += block_row_stride) {
uint2 rowcol = make_uint2(block_row_start, block_col_start);
#pragma unroll
for (int n = 0; n < size<2>(tensor) / 2; ++n, ++rowcol.y) {
// if (cute::thread(32, 0)) { printf("m = %d, n = %d, row = %d, col = %d\n", m, n, int(rowcol.x), int(rowcol.y));}
uint4 random_uint4 = flash::philox(seed, reinterpret_cast<unsigned long long&>(rowcol), offset);
// if (cute::thread0()) { printf("philox = %u, %d, %d, %d\n", random_uint4.x, random_uint4.y, random_uint4.z, random_uint4.w);}
uint8_t (&rnd_8)[16] = reinterpret_cast<uint8_t (&)[16]>(random_uint4);
// Special implementation for 16-bit types: we duplicate the threshold to the
// low and high 16 bits of a 32-bit value, then use the f16x2 comparison instruction
// to get a mask. The low 16 bits of the mask will be either 0xffff or 0x0000,
// and the high 16 bits will be either 0xffff or 0x0000, depending on whether
// the random value is less than the threshold.
// We then do a bit-wise AND between the mask and the original value (in 32-bit).
// We're exploiting the fact that floating point comparison is equivalent to integer
// comparison, since we're comparing unsigned integers whose top 8-bits are zero.
if (!encode_dropout_in_sign_bit
&& (std::is_same<T, cutlass::half_t>::value || std::is_same<T, cutlass::bfloat16_t>::value)) {
uint16_t rnd_16[16];
#pragma unroll
for (int i = 0; i < 16; i++) { rnd_16[i] = uint16_t(rnd_8[i]); }
uint32_t (&rnd_32)[8] = reinterpret_cast<uint32_t (&)[8]>(rnd_16);
#pragma unroll
for (int j = 0; j < 2; j++) {
Tensor tensor_uint32 = recast<uint32_t>(tensor(_, m, n * 2 + j));
// if (cute::thread0()) { printf("random = 0x%x, 0x%x, 0x%x, 0x%x\n", rnd_32[j * 4 + 0], rnd_32[j * 4 + 1], rnd_32[j * 4 + 2], rnd_32[j * 4 + 3]); }
// if (cute::thread0()) { printf("tensor_uint32 = 0x%x, 0x%x, 0x%x, 0x%x\n", tensor_uint32(0), tensor_uint32(1), tensor_uint32(2), tensor_uint32(3)); }
#pragma unroll
for (int i = 0; i < 4; i++) {
uint32_t mask;
asm volatile("set.le.u32.f16x2 %0, %1, %2;\n" : "=r"(mask) : "r"(rnd_32[j * 4 + i]), "r"(p_dropout_8bit_in_uint32_t));
tensor_uint32(i) &= mask;
}
// if (cute::thread0()) { printf("tensor_uint32 = 0x%x, 0x%x, 0x%x, 0x%x\n", tensor_uint32(0), tensor_uint32(1), tensor_uint32(2), tensor_uint32(3)); }
}
} else {
#pragma unroll
for (int j = 0; j < 2; j++) {
#pragma unroll
for (int i = 0; i < 8; i++) {
tensor(i, m, n * 2 + j) = encode_dropout(rnd_8[j * 8 + i] <= p_dropout_in_uint8_t, tensor(i, m, n * 2 + j));
}
Tensor tensor_uint32 = recast<uint32_t>(tensor(_, m, n * 2 + j));
// if (cute::thread0()) { printf("tensor_uint32 = 0x%x, 0x%x, 0x%x, 0x%x\n", tensor_uint32(0), tensor_uint32(1), tensor_uint32(2), tensor_uint32(3)); }
}
}
// // if ((threadIdx.x == 0) && (blockIdx.x == 0) && (blockIdx.y == 0)) {
// // printf("n = %d, ph Philox: %u, %u, %u, %u\n", n, rnd_8.x, rnd_8.y, rnd_8.z, rnd_8.w);
// // }
}
}
}
} // namespace flash