Skip to content
体验新版
项目
组织
正在加载...
登录
切换导航
打开侧边栏
Crayon鑫
Paddle
提交
a208dd64
P
Paddle
项目概览
Crayon鑫
/
Paddle
与 Fork 源项目一致
Fork自
PaddlePaddle / Paddle
通知
1
Star
1
Fork
0
代码
文件
提交
分支
Tags
贡献者
分支图
Diff
Issue
1
列表
看板
标记
里程碑
合并请求
0
Wiki
0
Wiki
分析
仓库
DevOps
项目成员
Pages
P
Paddle
项目概览
项目概览
详情
发布
仓库
仓库
文件
提交
分支
标签
贡献者
分支图
比较
Issue
1
Issue
1
列表
看板
标记
里程碑
合并请求
0
合并请求
0
Pages
分析
分析
仓库分析
DevOps
Wiki
0
Wiki
成员
成员
收起侧边栏
关闭侧边栏
动态
分支图
创建新Issue
提交
Issue看板
提交
a208dd64
编写于
10月 30, 2017
作者:
K
Kexin Zhao
浏览文件
操作
浏览文件
下载
电子邮件补丁
差异文件
add float16 data type
上级
e4733224
变更
1
隐藏空白更改
内联
并排
Showing
1 changed file
with
320 addition
and
45 deletion
+320
-45
paddle/math/float16.h
paddle/math/float16.h
+320
-45
未找到文件。
paddle/math/float16.h
浏览文件 @
a208dd64
...
@@ -18,7 +18,21 @@ limitations under the License. */
...
@@ -18,7 +18,21 @@ limitations under the License. */
#include <istream>
#include <istream>
#include <ostream>
#include <ostream>
#include <cuda.h>
#include <cuda.h> // seems need to delete it
#ifdef USE_EIGEN // delete this #if macro
#include "Eigen/src/Core/arch/CUDA/Half.h"
#endif
#ifdef __CUDACC__
#define PADDLE_HOSTDEVICE __host__ __device__
#if CUDA_VERSION >= 7050
#define PADDLE_CUDA_FP16
#include <cuda_fp16.h>
#endif // CUDA_VERSION >= 7050
#else
#define PADDLE_HOSTDEVICE
#endif // __CUDA_ARCH__
#ifdef __arm__
#ifdef __arm__
#define PADDLE_ARM_32
#define PADDLE_ARM_32
...
@@ -44,15 +58,9 @@ limitations under the License. */
...
@@ -44,15 +58,9 @@ limitations under the License. */
#define PADDLE_NEON_64
#define PADDLE_NEON_64
#endif
#endif
#ifdef __CUDA_ARCH__ // use __CUDACC__ instead
#if defined(PADDLE_ARM) && defined(PADDLE_NEON)
#define PADDLE_HOSTDEVICE __host__ __device__
#include <arm_neon.h>
#if CUDA_VERSION >= 7050
#endif
#define PADDLE_CUDA_FP16
#include <cuda_fp16.h>
#endif // CUDA_VERSION >= 7050
#else
#define PADDLE_HOSTDEVICE
#endif // __CUDA_ARCH__
#if !defined(__ANDROID__) && !defined(__APPLE__) && !defined(PADDLE_ARM)
#if !defined(__ANDROID__) && !defined(__APPLE__) && !defined(PADDLE_ARM)
#include <immintrin.h>
#include <immintrin.h>
...
@@ -62,7 +70,7 @@ limitations under the License. */
...
@@ -62,7 +70,7 @@ limitations under the License. */
#endif
#endif
#endif
#endif
#define PADDLE_ALIGN
ED
(x) __attribute__((aligned(x)))
#define PADDLE_ALIGN(x) __attribute__((aligned(x)))
// https://github.com/pytorch/pytorch/blob/master/torch/lib/ATen/Half.h
// https://github.com/pytorch/pytorch/blob/master/torch/lib/ATen/Half.h
template
<
typename
To
,
typename
From
>
template
<
typename
To
,
typename
From
>
...
@@ -72,70 +80,337 @@ To convert(From f) {
...
@@ -72,70 +80,337 @@ To convert(From f) {
namespace
paddle
{
namespace
paddle
{
class
float16
;
struct
float16
;
namespace
fp16_impl
{
// convert from float to half precision in round-to-nearest-even mode
// convert from float to half precision in round-to-nearest-even mode
float16
float2half_rn
(
float
f
);
PADDLE_HOSTDEVICE
inline
float16
float_to_half_rn
(
float
f
);
float
half2float
(
float16
h
);
PADDLE_HOSTDEVICE
inline
float
half_to_float
(
float16
h
);
PADDLE_HOSTDEVICE
inline
float16
uint16_to_half
(
uint16_t
x
);
}
// namespace fp16_impl
class
float16
{
// Use PADDLE_ALIGNED(2) to ensure that each float16 will be allocated
public:
// and aligned at least on a 2-byte boundary, which leads to efficient
uint16_t
val_
;
// memory access of float16 struct and also makes float16 compatible
// with CUDA half and Eigen::half data types.
struct
PADDLE_ALIGN
(
2
)
float16
{
uint16_t
x
;
PADDLE_HOSTDEVICE
inline
explicit
float16
()
:
x
(
0
)
{}
// explicit for different types, implicit for half and Eigen::half
PADDLE_HOSTDEVICE
inline
float16
()
{}
PADDLE_HOSTDEVICE
inline
float16
(
const
float16
&
h
)
:
x
(
h
.
x
)
{}
#ifdef PADDLE_CUDA_FP16
PADDLE_HOSTDEVICE
inline
float16
(
const
half
h
)
{
#if CUDA_VERSION >= 9000
x
=
reinterpret_cast
<
__half_raw
*>
(
&
h
)
->
x
;
#else
x
=
h
.
x
;
#endif // CUDA_VERSION >= 9000
}
#endif // PADDLE_CUDA_FP16
/*
#ifdef PADDLE_CUDA_FP16
#if CUDA_VERSION < 9000
PADDLE_HOSTDEVICE inline float16(const half& h) : x(h.x) {}
#else
PADDLE_HOSTDEVICE inline float16(const __half_raw& h) : x(h.x) {}
PADDLE_HOSTDEVICE inline float16(const half& h)
: x(*reinterpret_cast<uint16_t*>(&h)) {}
#endif // CUDA_VERSION < 9000
#endif // PADDLE_CUDA_FP16
*/
#ifdef USE_EIGEN
PADDLE_HOSTDEVICE
inline
float16
(
const
Eigen
::
half
&
h
)
:
x
(
h
.
x
)
{}
#endif // USE_EIGEN
#if defined(PADDLE_ARM) && defined(PADDLE_NEON)
// __fp16 is a native half precision data type for arm cpu,
// float16_t is an alias for __fp16 in arm_fp16.h
// which is included in arm_neon.h
PADDLE_HOSTDEVICE
inline
float16
(
const
float16_t
h
)
{
x
=
*
reinterpret_cast
<
uint16_t
*>
(
&
h
);
}
#endif
PADDLE_HOSTDEVICE
inline
explicit
float16
(
bool
b
)
:
x
(
b
?
0x3c00
:
0
)
{}
PADDLE_HOSTDEVICE
inline
explicit
float16
(
float
val
)
{
PADDLE_HOSTDEVICE
inline
explicit
float16
(
float
val
)
{
float16
res
=
float2half_rn
(
val
);
float16
res
=
fp16_impl
::
float_to_half_rn
(
val
);
x
=
res
.
x
;
}
template
<
class
T
>
PADDLE_HOSTDEVICE
inline
explicit
float16
(
const
T
&
val
)
{
float16
res
=
fp16_impl
::
float_to_half_rn
(
static_cast
<
float
>
(
val
));
x
=
res
.
x
;
x
=
res
.
x
;
}
}
PADDLE_HOSTDEVICE
inline
float16
&
operator
=
(
const
float16
&
rhs
)
{
x
=
rhs
.
x
;
return
*
this
;
}
#ifdef PADDLE_CUDA_FP16
PADDLE_HOSTDEVICE
inline
float16
&
operator
=
(
const
half
rhs
)
{
#if CUDA_VERSION >= 9000
x
=
reinterpret_cast
<
__half_raw
*>
(
&
rhs
)
->
x
;
#else
x
=
rhs
.
x
;
#endif
return
*
this
;
}
#endif
#ifdef USE_EIGEN
PADDLE_HOSTDEVICE
inline
float16
&
operator
=
(
const
Eigen
::
half
&
rhs
)
{
x
=
rhs
.
x
;
return
*
this
;
}
#endif // USE_EIGEN
#if defined(PADDLE_ARM) && defined(PADDLE_NEON)
PADDLE_HOSTDEVICE
inline
float16
&
operator
=
(
const
float16_t
rhs
)
{
x
=
*
reinterpret_cast
<
uint16_t
*>
(
&
rhs
);
return
*
this
;
}
#endif
/*
PADDLE_HOSTDEVICE inline explicit float16(int val) {
PADDLE_HOSTDEVICE inline explicit float16(int val) {
float16
res
=
cpu_float2
half_rn
(
static_cast
<
float
>
(
val
));
float16 res =
fp16_impl::float_to_
half_rn(static_cast<float>(val));
x = res.x;
x = res.x;
}
}
PADDLE_HOSTDEVICE inline explicit float16(double val) {
PADDLE_HOSTDEVICE inline explicit float16(double val) {
float16
res
=
cpu_float2
half_rn
(
static_cast
<
float
>
(
val
));
float16 res =
fp16_impl::float_to_
half_rn(static_cast<float>(val));
x = res.x;
x = res.x;
}
}
*/
#ifdef PADDLE_CUDA_FP16
PADDLE_HOSTDEVICE
inline
operator
half
()
{
#if CUDA_VERSION >= 9000
__half_raw
h
;
h
.
x
=
x
;
return
half
(
h
);
#else
half
h
;
h
.
x
=
x
;
return
h
;
#endif // CUDA_VERSION >= 9000
}
#endif // PADDLE_CUDA_FP16
// Use PADDLE_ALIGNED(2) to ensure that each float16 will be allocated
#ifdef USE_EIGEN
// and aligned at least on a 2-byte boundary, which leads to efficient
PADDLE_HOSTDEVICE
inline
operator
Eigen
::
half
()
{
// memory access of float16 struct.
Eigen
::
half
h
;
}
PADDLE_ALIGNED
(
2
);
h
.
x
=
x
;
return
h
;
}
#endif // USE_EIGEN
#if defined(PADDLE_ARM) && defined(PADDLE_NEON)
PADDLE_HOSTDEVICE
inline
operator
float16_t
()
{
float16
h
=
*
this
;
return
*
reinterpret_cast
<
float16_t
*>
(
&
h
);
}
#endif
PADDLE_HOSTDEVICE
inline
explicit
operator
bool
()
{
return
(
x
&
0x7fff
)
!=
0
;
}
PADDLE_HOSTDEVICE
inline
explicit
operator
int8_t
()
{
return
static_cat
<
int8_t
>
(
fp16_impl
::
half_to_float
(
*
this
));
}
PADDLE_HOSTDEVICE
inline
explicit
operator
uint8_t
()
{
return
static_cat
<
uint8_t
>
(
fp16_impl
::
half_to_float
(
*
this
));
}
PADDLE_HOSTDEVICE
inline
explicit
operator
int16_t
()
{
return
static_cat
<
int16_t
>
(
fp16_impl
::
half_to_float
(
*
this
));
}
PADDLE_HOSTDEVICE
inline
explicit
operator
uint16_t
()
{
return
static_cat
<
uint16_t
>
(
fp16_impl
::
half_to_float
(
*
this
));
}
PADDLE_HOSTDEVICE
inline
explicit
operator
int32_t
()
{
return
static_cat
<
int32_t
>
(
fp16_impl
::
half_to_float
(
*
this
));
}
PADDLE_HOSTDEVICE
inline
explicit
operator
uint32_t
()
{
return
static_cat
<
uint32_t
>
(
fp16_impl
::
half_to_float
(
*
this
));
}
PADDLE_HOSTDEVICE
inline
explicit
operator
int64_t
()
{
return
static_cat
<
int64_t
>
(
fp16_impl
::
half_to_float
(
*
this
));
}
PADDLE_HOSTDEVICE
inline
explicit
operator
uint64_t
()
{
return
static_cat
<
uint64_t
>
(
fp16_impl
::
half_to_float
(
*
this
));
}
PADDLE_HOSTDEVICE
inline
explicit
operator
float
()
{
return
fp16_impl
::
half_to_float
(
*
this
);
}
PADDLE_HOSTDEVICE
inline
explicit
operator
double
()
{
return
static_cat
<
double
>
(
fp16_impl
::
half_to_float
(
*
this
));
}
};
// arithmetic operators
#if defined(PADDLE_CUDA_FP16) && defined(__CUDA_ARCH__) && __CUDA_ARCH__ >= 530
__device__
inline
float16
operator
+
(
const
float16
&
a
,
const
float16
&
b
)
{
return
float16
(
__hadd
(
a
,
b
));
}
__device__
inline
float16
operator
-
(
const
float16
&
a
,
const
float16
&
b
)
{
return
__hsub
(
a
,
b
);
}
__device__
inline
float16
operator
*
(
const
float16
&
a
,
const
float16
&
b
)
{
return
__hmul
(
a
,
b
);
}
#elif // on arm cpu
#else
#endif
namespace
fp16_impl
{
namespace
fp16_impl
{
// Conversion routine adapted from
// http://stackoverflow.com/questions/1659440/32-bit-to-16-bit-floating-point-conversion
Union
Bits
{
Union
Bits
{
float
f
;
float
f
;
int32_t
si
;
int32_t
si
;
uint32_t
ui
;
uint32_t
ui
;
};
};
static
const
int
shift
=
13
;
const
int
shift
=
13
;
static
const
int
shiftSign
=
16
;
const
int
shiftSign
=
16
;
const
int32_t
infN
=
0x7F800000
;
const
int32_t
maxN
=
0x477FE000
;
// max flt16 as flt32
const
int32_t
minN
=
0x38800000
;
// min flt16 normal as flt32
const
int32_t
sigN
=
0x80000000
;
// sign bit
constexpr
int32_t
infC
=
infN
>>
shift
;
constexpr
int32_t
nanN
=
(
infC
+
1
)
<<
shift
;
// minimum flt16 nan as float32
constexpr
int32_t
maxC
=
maxN
>>
shift
;
constexpr
int32_t
minC
=
minN
>>
shift
;
constexpr
int32_t
sigC
=
sigN
>>
shiftSign
;
const
int32_t
mulN
=
0x52000000
;
//(1 << 23) / minN
const
int32_t
mulC
=
0x33800000
;
// minN / (1 << (23 - shift))
const
int32_t
subC
=
0x003FF
;
// max flt32 subnormal downshifted
const
int32_t
norC
=
0x00400
;
// min flt32 normal downshifted
constexpr
int32_t
maxD
=
infC
-
maxC
-
1
;
constexpr
int32_t
minD
=
minC
-
subC
-
1
;
PADDLE_HOSTDEVICE
inline
float16
float_to_half_rn
(
float
f
)
{
#if defined(PADDLE_CUDA_FP16) && defined(__CUDA_ARCH__) && __CUDA_ARCH__ >= 300
half
tmp
=
__float2half
(
f
);
return
*
reinterpret_cast
<
float16
*>
(
&
(
tmp
));
#elif defined(__F16C__)
float16
res
;
res
.
x
=
_cvtss_sh
(
f
,
0
);
return
res
;
#elif defined(PADDLE_ARM_64) // test on RPI
float16
res
;
asm
volatile
(
"ld1 {v0.s}[0], [%[float_ptr]]
\n
"
"FCVT h0, s0
\n
"
"st1 {v0.h}[0], [%[half_ptr]]
\n
"
:
// outputs
:
// inputs
[
float_ptr
]
"r"
(
&
f
),
[
half_ptr
]
"r"
(
&
(
res
.
x
))
:
// clobbers
"memory"
,
"v0"
);
return
res
;
static
const
int32_t
infN
=
0x7F800000
;
#else
static
const
int32_t
maxN
=
0x477FE000
;
// max flt16 as flt32
// Conversion routine adapted from
static
const
int32_t
minN
=
0x38800000
;
// min flt16 normal as flt32
// http://stackoverflow.com/questions/1659440/32-bit-to-16-bit-floating-point-conversion
static
const
int32_t
sigN
=
0x80000000
;
// sign bit
Bits
v
,
s
;
v
.
f
=
f
;
uint32_t
sign
=
v
.
si
&
sigN
;
v
.
si
^=
sign
;
sign
>>=
shiftSign
;
// logical shift
s
.
si
=
mulN
;
s
.
si
=
s
.
f
*
v
.
f
;
// correct subnormals
v
.
si
^=
(
s
.
si
^
v
.
si
)
&
-
(
minN
>
v
.
si
);
v
.
si
^=
(
infN
^
v
.
si
)
&
-
((
infN
>
v
.
si
)
&
(
v
.
si
>
maxN
));
v
.
si
^=
(
nanN
^
v
.
si
)
&
-
((
nanN
>
v
.
si
)
&
(
v
.
si
>
infN
));
v
.
ui
>>=
shift
;
// logical shift
v
.
si
^=
((
v
.
si
-
maxD
)
^
v
.
si
)
&
-
(
v
.
si
>
maxC
);
v
.
si
^=
((
v
.
si
-
minD
)
^
v
.
si
)
&
-
(
v
.
si
>
subC
);
float16
res
;
res
.
x
=
v
.
ui
|
sign
;
return
res
;
static
constexpr
int32_t
infC
=
infN
>>
shift
;
#endif
static
constexpr
int32_t
nanN
=
(
infC
+
1
)
}
<<
shift
;
// minimum flt16 nan as float32
static
constexpr
int32_t
maxC
=
maxN
>>
shift
;
static
constexpr
int32_t
minC
=
minN
>>
shift
;
static
constexpr
int32_t
sigC
=
sigN
>>
shiftSign
;
static
const
int32_t
mulN
=
0x52000000
;
//(1 << 23) / minN
PADDLE_HOSTDEVICE
inline
float
half_to_float
(
float16
h
)
{
static
const
int32_t
mulC
=
0x33800000
;
// minN / (1 << (23 - shift))
#if defined(PADDLE_CUDA_FP16) && defined(__CUDA_ARCH__) && __CUDA_ARCH__ >= 300
static
const
int32_t
subC
=
0x003FF
;
// max flt32 subnormal downshifted
half
tmp
=
*
reinterpret_cast
<
half
*>
(
&
h
);
static
const
int32_t
norC
=
0x00400
;
// min flt32 normal downshifted
return
__half2float
(
h
);
#elif defined(__F16C__)
return
_cvtsh_ss
(
h
.
x
);
#elif defined(PADDLE_ARM_64) // test on RPI
float
res
;
asm
volatile
(
"ld1 {v0.h}[0], [%[half_ptr]]
\n
"
"FCVT s0, h0
\n
"
"st1 {v0.s}[0], [%[float_ptr]]
\n
"
:
// outputs
:
// inputs
[
half_ptr
]
"r"
(
&
(
h
.
x
)),
[
float_ptr
]
"r"
(
&
res
)
:
// clobbers
"memory"
,
"v0"
);
return
res
;
static
constexpr
int32_t
maxD
=
infC
-
maxC
-
1
;
#else
static
constexpr
int32_t
minD
=
minC
-
subC
-
1
;
// Conversion routine adapted from
// http://stackoverflow.com/questions/1659440/32-bit-to-16-bit-floating-point-conversion
Bits
v
;
v
.
ui
=
x
;
int32_t
sign
=
v
.
si
&
sigC
;
v
.
si
^=
sign
;
sign
<<=
shiftSign
;
v
.
si
^=
((
v
.
si
+
minD
)
^
v
.
si
)
&
-
(
v
.
si
>
subC
);
v
.
si
^=
((
v
.
si
+
maxD
)
^
v
.
si
)
&
-
(
v
.
si
>
maxC
);
Bits
s
;
s
.
si
=
mulC
;
s
.
f
*=
v
.
si
;
int32_t
mask
=
-
(
norC
>
v
.
si
);
v
.
si
<<=
shift
;
v
.
si
^=
(
s
.
si
^
v
.
si
)
&
mask
;
v
.
si
|=
sign
;
return
v
.
f
;
#endif
}
PADDLE_HOSTDEVICE
inline
float16
uint16_to_half
(
uint16_t
x
)
{
float16
res
;
res
.
x
=
x
;
return
res
;
}
}
// namespace half_impl
}
// namespace half_impl
...
...
编辑
预览
Markdown
is supported
0%
请重试
或
添加新附件
.
添加附件
取消
You are about to add
0
people
to the discussion. Proceed with caution.
先完成此消息的编辑!
取消
想要评论请
注册
或
登录