mirror of https://github.com/hpcaitech/ColossalAI
[NFC] polish colossalai/kernel/cuda_native/csrc/multi_tensor_adam.cu code style (#667)
encmps
binmakeswell
parent
e4bcff9b0f
commit
79ccfa4310
|
|
@ -1,4 +1,5 @@
|
|||
// modified from https://github.com/NVIDIA/apex/blob/master/csrc/multi_tensor_adam.cu
|
||||
// modified from
|
||||
// https://github.com/NVIDIA/apex/blob/master/csrc/multi_tensor_adam.cu
|
||||
#include <ATen/ATen.h>
|
||||
#include <ATen/AccumulateType.h>
|
||||
#include <ATen/cuda/CUDAContext.h>
|
||||
|
|
@ -8,14 +9,13 @@
|
|||
|
||||
#include <assert.h>
|
||||
|
||||
#include "type_shim.h"
|
||||
#include "multi_tensor_apply.cuh"
|
||||
#include "type_shim.h"
|
||||
|
||||
#define BLOCK_SIZE 512
|
||||
#define ILP 4
|
||||
|
||||
typedef enum
|
||||
{
|
||||
typedef enum {
|
||||
ADAM_MODE_0 = 0, // L2 regularization mode
|
||||
ADAM_MODE_1 = 1 // Decoupled weight decay mode(AdamW)
|
||||
} adamMode_t;
|
||||
|
|
@ -23,21 +23,12 @@ typedef enum
|
|||
using MATH_T = float;
|
||||
|
||||
template <typename T_g, typename T_p>
|
||||
struct AdamFunctor
|
||||
{
|
||||
struct AdamFunctor {
|
||||
__device__ __forceinline__ void operator()(
|
||||
int chunk_size,
|
||||
volatile int *noop_gmem,
|
||||
TensorListMetadata<4> &tl,
|
||||
const float beta1,
|
||||
const float beta2,
|
||||
const float beta1_correction,
|
||||
const float beta2_correction,
|
||||
const float epsilon,
|
||||
const float lr,
|
||||
adamMode_t mode,
|
||||
const float decay)
|
||||
{
|
||||
int chunk_size, volatile int *noop_gmem, TensorListMetadata<4> &tl,
|
||||
const float beta1, const float beta2, const float beta1_correction,
|
||||
const float beta2_correction, const float epsilon, const float lr,
|
||||
adamMode_t mode, const float decay) {
|
||||
// I'd like this kernel to propagate infs/nans.
|
||||
// if(*noop_gmem == 1)
|
||||
// return;
|
||||
|
|
@ -65,27 +56,21 @@ struct AdamFunctor
|
|||
n -= chunk_idx * chunk_size;
|
||||
|
||||
// see note in multi_tensor_scale_kernel.cu
|
||||
for (int i_start = 0;
|
||||
i_start < n && i_start < chunk_size;
|
||||
i_start += blockDim.x * ILP)
|
||||
{
|
||||
for (int i_start = 0; i_start < n && i_start < chunk_size;
|
||||
i_start += blockDim.x * ILP) {
|
||||
MATH_T r_g[ILP];
|
||||
MATH_T r_p[ILP];
|
||||
MATH_T r_m[ILP];
|
||||
MATH_T r_v[ILP];
|
||||
#pragma unroll
|
||||
for (int ii = 0; ii < ILP; ii++)
|
||||
{
|
||||
for (int ii = 0; ii < ILP; ii++) {
|
||||
int i = i_start + threadIdx.x + ii * blockDim.x;
|
||||
if (i < n && i < chunk_size)
|
||||
{
|
||||
if (i < n && i < chunk_size) {
|
||||
r_g[ii] = g[i];
|
||||
r_p[ii] = p[i];
|
||||
r_m[ii] = m[i];
|
||||
r_v[ii] = v[i];
|
||||
}
|
||||
else
|
||||
{
|
||||
} else {
|
||||
r_g[ii] = MATH_T(0);
|
||||
r_p[ii] = MATH_T(0);
|
||||
r_m[ii] = MATH_T(0);
|
||||
|
|
@ -93,10 +78,8 @@ struct AdamFunctor
|
|||
}
|
||||
}
|
||||
#pragma unroll
|
||||
for (int ii = 0; ii < ILP; ii++)
|
||||
{
|
||||
if (mode == ADAM_MODE_0)
|
||||
{ // L2
|
||||
for (int ii = 0; ii < ILP; ii++) {
|
||||
if (mode == ADAM_MODE_0) { // L2
|
||||
r_g[ii] = r_g[ii] + (decay * r_p[ii]);
|
||||
r_m[ii] = beta1 * r_m[ii] + (1 - beta1) * r_g[ii];
|
||||
r_v[ii] = beta2 * r_v[ii] + (1 - beta2) * r_g[ii] * r_g[ii];
|
||||
|
|
@ -105,9 +88,7 @@ struct AdamFunctor
|
|||
MATH_T denom = sqrtf(next_v_unbiased) + epsilon;
|
||||
MATH_T update = next_m_unbiased / denom;
|
||||
r_p[ii] = r_p[ii] - (lr * update);
|
||||
}
|
||||
else
|
||||
{ // weight decay
|
||||
} else { // weight decay
|
||||
r_m[ii] = beta1 * r_m[ii] + (1 - beta1) * r_g[ii];
|
||||
r_v[ii] = beta2 * r_v[ii] + (1 - beta2) * r_g[ii] * r_g[ii];
|
||||
MATH_T next_m_unbiased = r_m[ii] / beta1_correction;
|
||||
|
|
@ -118,11 +99,9 @@ struct AdamFunctor
|
|||
}
|
||||
}
|
||||
#pragma unroll
|
||||
for (int ii = 0; ii < ILP; ii++)
|
||||
{
|
||||
for (int ii = 0; ii < ILP; ii++) {
|
||||
int i = i_start + threadIdx.x + ii * blockDim.x;
|
||||
if (i < n && i < chunk_size)
|
||||
{
|
||||
if (i < n && i < chunk_size) {
|
||||
p[i] = r_p[ii];
|
||||
m[i] = r_m[ii];
|
||||
v[i] = r_v[ii];
|
||||
|
|
@ -132,46 +111,29 @@ struct AdamFunctor
|
|||
}
|
||||
};
|
||||
|
||||
void multi_tensor_adam_cuda(
|
||||
int chunk_size,
|
||||
at::Tensor noop_flag,
|
||||
void multi_tensor_adam_cuda(int chunk_size, at::Tensor noop_flag,
|
||||
std::vector<std::vector<at::Tensor>> tensor_lists,
|
||||
const float lr,
|
||||
const float beta1,
|
||||
const float beta2,
|
||||
const float epsilon,
|
||||
const int step,
|
||||
const int mode,
|
||||
const float lr, const float beta1,
|
||||
const float beta2, const float epsilon,
|
||||
const int step, const int mode,
|
||||
const int bias_correction,
|
||||
const float weight_decay)
|
||||
{
|
||||
const float weight_decay) {
|
||||
using namespace at;
|
||||
|
||||
// Handle bias correction mode
|
||||
float bias_correction1 = 1.0f, bias_correction2 = 1.0f;
|
||||
if (bias_correction == 1)
|
||||
{
|
||||
if (bias_correction == 1) {
|
||||
bias_correction1 = 1 - std::pow(beta1, step);
|
||||
bias_correction2 = 1 - std::pow(beta2, step);
|
||||
}
|
||||
|
||||
DISPATCH_FLOAT_AND_HALF_FOR_G_P(
|
||||
tensor_lists[0][0].scalar_type(),
|
||||
tensor_lists[1][0].scalar_type(), 0, "adam",
|
||||
multi_tensor_apply<4>(
|
||||
BLOCK_SIZE,
|
||||
chunk_size,
|
||||
noop_flag,
|
||||
tensor_lists,
|
||||
AdamFunctor<g_scalar_t_0, p_scalar_t_0>(),
|
||||
beta1,
|
||||
beta2,
|
||||
bias_correction1,
|
||||
bias_correction2,
|
||||
epsilon,
|
||||
lr,
|
||||
(adamMode_t)mode,
|
||||
weight_decay);)
|
||||
tensor_lists[0][0].scalar_type(), tensor_lists[1][0].scalar_type(), 0,
|
||||
"adam",
|
||||
multi_tensor_apply<4>(BLOCK_SIZE, chunk_size, noop_flag, tensor_lists,
|
||||
AdamFunctor<g_scalar_t_0, p_scalar_t_0>(), beta1,
|
||||
beta2, bias_correction1, bias_correction2, epsilon,
|
||||
lr, (adamMode_t)mode, weight_decay);)
|
||||
|
||||
AT_CUDA_CHECK(cudaGetLastError());
|
||||
}
|
||||
Loading…
Reference in New Issue