Skip to content

GitLab

Commit a298c926 authored by Umang Yadav's avatar Umang Yadav
Browse files

deprecate pytorch implementation

parent 40c2df86
Hide whitespace changes
Inline Side-by-side
Showing with 0 additions and 525 deletions
src/include/migraphx/fp8e4m3fnuz.hpp deleted 100644 → 0
View file @ 40c2df86
/*
* The MIT License (MIT)
*
* Copyright (c) 2015-2023 Advanced Micro Devices, Inc. All rights reserved.
*
* Permission is hereby granted, free of charge, to any person obtaining a copy
* of this software and associated documentation files (the "Software"), to deal
* in the Software without restriction, including without limitation the rights
* to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
* copies of the Software, and to permit persons to whom the Software is
* furnished to do so, subject to the following conditions:
*
* The above copyright notice and this permission notice shall be included in
* all copies or substantial portions of the Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
* AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
* OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
* THE SOFTWARE.
*/
#ifndef MIGRAPHX_GUARD_RTGLIB_FP8E4M3FNUZ_HPP
#define MIGRAPHX_GUARD_RTGLIB_FP8E4M3FNUZ_HPP
/// Defines the Float8_e4m3fnuz type (8-bit floating-point) including
/// conversions to standard C types and basic arithmetic operations. Note that
/// arithmetic operations are implemented by converting to floating point and
/// performing the operation in float32.
///
/// Binary configuration remains the same as Float8_e4m3fn:
/// s eeee mmm
/// 1 sign bit
/// 4 exponent bits
/// 3 mantissa bits
///
/// The key differences versus Float8_e4m3fn are:
/// bias = 8
/// no infinities or negative zero
/// NaN only when sign bit is 1, rest all 0s
///
/// Implementation based on the paper https://arxiv.org/pdf/2206.02915.pdf and
/// the existing Float8_e4m3fn implementation.
#include <type_traits>
#include <cmath>
#include <cstdint>
#include <climits>
#include <cstring>
#include <iosfwd>
#include <limits>
#include <sstream>
#include <iostream>
#include <migraphx/config.hpp>
#include <string>
#include <utility>
#if defined(__HIP_PLATFORM_AMD__) || defined(__HIP_PLATFORM_HCC__)
// MIGraphX by default does not have device code in the regular compilation paths,
// therefore, when this file is used from the host side, compilation takes much
// longer. By guarding the __device__ directive we can control that such compilation
// only happens for kernels which include this file.
// need to include hip_runtime.h otherwise it complains about __host__ and __device__
#include <hip/hip_runtime.h>
#define MIGRAPHX_HIP_HOST_DEVICE __host__ __device__
#else
#define MIGRAPHX_HIP_HOST_DEVICE
#endif
#pragma clang diagnostic push
#pragma clang diagnostic ignored "-Wimplicit-int-float-conversion"
#pragma clang diagnostic ignored "-Wold-style-cast"
#pragma clang diagnostic ignored "-Wreserved-identifier"
#pragma clang diagnostic ignored "-Wfloat-equal"
#pragma clang diagnostic ignored "-Wabsolute-value"
namespace migraphx {
inline namespace MIGRAPHX_INLINE_NS {
namespace detail {
inline MIGRAPHX_HIP_HOST_DEVICE float fp32_from_bits(uint32_t w)
{
union
{
uint32_t as_bits;
float as_value;
} fp32 = {w};
return fp32.as_value;
}
inline MIGRAPHX_HIP_HOST_DEVICE uint32_t fp32_to_bits(float f)
{
union
{
float as_value;
uint32_t as_bits;
} fp32 = {f};
return fp32.as_bits;
}
/*
* Convert a 8-bit floating-point number in fp8 E4M3FNUZ format, in bit
* representation, to a 32-bit floating-point number in IEEE single-precision
* format, in bit representation.
*
* @note The implementation doesn't use any floating-point operations.
*/
inline MIGRAPHX_HIP_HOST_DEVICE constexpr float fp8e4m3fnuz_to_fp32_value(uint8_t input)
{
constexpr float e4m3fnuz_lut[256] = {
0.0f, 0.0009765625f, 0.001953125f,
0.0029296875f, 0.00390625f, 0.0048828125f,
0.005859375f, 0.0068359375f, 0.0078125f,
0.0087890625f, 0.009765625f, 0.0107421875f,
0.01171875f, 0.0126953125f, 0.013671875f,
0.0146484375f, 0.015625f, 0.017578125f,
0.01953125f, 0.021484375f, 0.0234375f,
0.025390625f, 0.02734375f, 0.029296875f,
0.03125f, 0.03515625f, 0.0390625f,
0.04296875f, 0.046875f, 0.05078125f,
0.0546875f, 0.05859375f, 0.0625f,
0.0703125f, 0.078125f, 0.0859375f,
0.09375f, 0.1015625f, 0.109375f,
0.1171875f, 0.125f, 0.140625f,
0.15625f, 0.171875f, 0.1875f,
0.203125f, 0.21875f, 0.234375f,
0.25f, 0.28125f, 0.3125f,
0.34375f, 0.375f, 0.40625f,
0.4375f, 0.46875f, 0.5f,
0.5625f, 0.625f, 0.6875f,
0.75f, 0.8125f, 0.875f,
0.9375f, 1.0f, 1.125f,
1.25f, 1.375f, 1.5f,
1.625f, 1.75f, 1.875f,
2.0f, 2.25f, 2.5f,
2.75f, 3.0f, 3.25f,
3.5f, 3.75f, 4.0f,
4.5f, 5.0f, 5.5f,
6.0f, 6.5f, 7.0f,
7.5f, 8.0f, 9.0f,
10.0f, 11.0f, 12.0f,
13.0f, 14.0f, 15.0f,
16.0f, 18.0f, 20.0f,
22.0f, 24.0f, 26.0f,
28.0f, 30.0f, 32.0f,
36.0f, 40.0f, 44.0f,
48.0f, 52.0f, 56.0f,
60.0f, 64.0f, 72.0f,
80.0f, 88.0f, 96.0f,
104.0f, 112.0f, 120.0f,
128.0f, 144.0f, 160.0f,
176.0f, 192.0f, 208.0f,
224.0f, 240.0f, std::numeric_limits<float>::quiet_NaN(),
-0.0009765625f, -0.001953125f, -0.0029296875f,
-0.00390625f, -0.0048828125f, -0.005859375f,
-0.0068359375f, -0.0078125f, -0.0087890625f,
-0.009765625f, -0.0107421875f, -0.01171875f,
-0.0126953125f, -0.013671875f, -0.0146484375f,
-0.015625f, -0.017578125f, -0.01953125f,
-0.021484375f, -0.0234375f, -0.025390625f,
-0.02734375f, -0.029296875f, -0.03125f,
-0.03515625f, -0.0390625f, -0.04296875f,
-0.046875f, -0.05078125f, -0.0546875f,
-0.05859375f, -0.0625f, -0.0703125f,
-0.078125f, -0.0859375f, -0.09375f,
-0.1015625f, -0.109375f, -0.1171875f,
-0.125f, -0.140625f, -0.15625f,
-0.171875f, -0.1875f, -0.203125f,
-0.21875f, -0.234375f, -0.25f,
-0.28125f, -0.3125f, -0.34375f,
-0.375f, -0.40625f, -0.4375f,
-0.46875f, -0.5f, -0.5625f,
-0.625f, -0.6875f, -0.75f,
-0.8125f, -0.875f, -0.9375f,
-1.0f, -1.125f, -1.25f,
-1.375f, -1.5f, -1.625f,
-1.75f, -1.875f, -2.0f,
-2.25f, -2.5f, -2.75f,
-3.0f, -3.25f, -3.5f,
-3.75f, -4.0f, -4.5f,
-5.0f, -5.5f, -6.0f,
-6.5f, -7.0f, -7.5f,
-8.0f, -9.0f, -10.0f,
-11.0f, -12.0f, -13.0f,
-14.0f, -15.0f, -16.0f,
-18.0f, -20.0f, -22.0f,
-24.0f, -26.0f, -28.0f,
-30.0f, -32.0f, -36.0f,
-40.0f, -44.0f, -48.0f,
-52.0f, -56.0f, -60.0f,
-64.0f, -72.0f, -80.0f,
-88.0f, -96.0f, -104.0f,
-112.0f, -120.0f, -128.0f,
-144.0f, -160.0f, -176.0f,
-192.0f, -208.0f, -224.0f,
-240.0f,
};
return e4m3fnuz_lut[input];
}
/*
* Convert a 32-bit floating-point number in IEEE single-precision format to a
* 8-bit floating-point number in fp8 E4M3FNUZ format, in bit representation.
*/
inline MIGRAPHX_HIP_HOST_DEVICE uint8_t fp8e4m3fnuz_from_fp32_value(float f)
{
/*
* Binary representation of 256.0f, which is the first value not representable
* (i.e. the first value which would overflow in to the sign bit, resulting in
* a NaN) in fp8e4m3fnuz range:
* 1 0000 000 - fp8e4m3fnuz
* 0 10000110 00000000000000000000000 - fp32
*/
constexpr uint32_t fnuz_max = UINT32_C(0x87) << 23;
/*
* A mask for converting fp32 numbers lower than fp8e4m3fnuz normal range
* into denormalized representation.
* magic number: ((127 - 8) + (23 - 3) + 1)
*/
constexpr uint32_t denorm_mask = UINT32_C(0x8C) << 23;
uint32_t f_bits = fp32_to_bits(f);
uint32_t result = 0u;
/*
* Extract the sign of the input number into the high bit of the 32-bit word:
*
* +---+----------------------------------+
* | S |0000000 00000000 00000000 00000000|
* +---+----------------------------------+
* Bits 31 0-31
*/
const uint32_t sign = f_bits & UINT32_C(0x80000000);
/*
* Set sign bit to 0
*/
f_bits ^= sign;
if(f_bits >= fnuz_max)
{
// NaN -- sign bit set to 1, rest 0s
return 0x80;
}
if(f_bits < (UINT32_C(121) << 23))
{
// Input number is smaller than 2^(-7), which is the smallest
// fp8e4m3fnuz normal number
f_bits = fp32_to_bits(fp32_from_bits(f_bits) + fp32_from_bits(denorm_mask));
result = static_cast<uint8_t>(f_bits - denorm_mask);
if(result == 0)
{
// fnuz types don't have negative zero.
return 0;
}
}
else
{
// resulting mantissa is odd
uint8_t mant_odd = (f_bits >> 20) & 1;
// update exponent, rounding bias part 1
f_bits += ((uint32_t)(8 - 127) << 23) + 0x7FFFF;
// rounding bias part 2
f_bits += mant_odd;
// take the bits!
result = static_cast<uint8_t>(f_bits >> 20);
}
result |= sign >> 24;
return result;
}
struct expr
{
/// Conversion constructor.
/// \param f single-precision value to convert
explicit constexpr expr(float f) noexcept : value_(f) {}
/// Conversion to single-precision.
/// \return single precision value representing expression value
constexpr operator float() const noexcept { return value_; }
private:
/// Internal expression value stored in single-precision.
float value_;
};
} // namespace detail
/*
overloads using migraphx::fp8e4m3fnuz may not be necessary since they can be implicitly casted to
float that is how half.hpp is implementing it.
this operators can't be friend since it leads to conflicting candidates with inbuilt operators (due
to implict cast to other types probably)
*/
// NOLINTNEXTLINE
#define MIGRAPHX_FP8_UNARY_OP(unary_op) \
constexpr migraphx::fp8e4m3fnuz& MIGRAPHX_HIP_HOST_DEVICE operator unary_op( \
const migraphx::fp8e4m3fnuz& rhs) \
{ \
float y = float(data_); \
y unary_op float(rhs); \
data_ = detail::fp8e4m3fnuz_from_fp32_value(y); \
return *this; \
} \
constexpr migraphx::fp8e4m3fnuz& MIGRAPHX_HIP_HOST_DEVICE operator unary_op(const float& rhs) \
{ \
float y = float(data_); \
y unary_op rhs; \
data_ = detail::fp8e4m3fnuz_from_fp32_value(y); \
return *this; \
}
// NOLINTNEXTLINE
#define MIGRAPHX_FP8_BINARY_OP(binary_op, T) \
friend constexpr T MIGRAPHX_HIP_HOST_DEVICE operator binary_op( \
const migraphx::fp8e4m3fnuz& lhs, const migraphx::fp8e4m3fnuz& rhs) \
{ \
return T(float(lhs) binary_op float(rhs)); \
}
// NOLINTNEXTLINE
#define MIGRAPHX_FP8_MATH(name, fname) \
migraphx::fp8e4m3fnuz MIGRAPHX_HIP_HOST_DEVICE name(migraphx::fp8e4m3fnuz x) \
{ \
return migraphx::fp8e4m3fnuz(fname(float(x))); \
}
} // namespace MIGRAPHX_INLINE_NS
struct alignas(1) fp8e4m3fnuz
{
uint8_t data_;
struct from_bits_t
{
};
static constexpr MIGRAPHX_HIP_HOST_DEVICE from_bits_t from_bits() { return from_bits_t(); }
MIGRAPHX_HIP_HOST_DEVICE fp8e4m3fnuz() : data_(0) {}
MIGRAPHX_HIP_HOST_DEVICE constexpr fp8e4m3fnuz(uint8_t bits, from_bits_t) : data_(bits) {}
MIGRAPHX_HIP_HOST_DEVICE fp8e4m3fnuz(const fp8e4m3fnuz& y) = default;
inline explicit constexpr MIGRAPHX_HIP_HOST_DEVICE fp8e4m3fnuz(float value)
: data_(detail::fp8e4m3fnuz_from_fp32_value(value))
{
}
#if !defined(__HIP_NO_F8_CONVERSIONS__)
// for the device kernels, this needs to be disabled since implicit_conversion op can type cast
// any type to any other type and that results in conflicts in candidate overload resolutions.
fp8e4m3fnuz& MIGRAPHX_HIP_HOST_DEVICE operator=(float rhs)
{
data_ = detail::fp8e4m3fnuz_from_fp32_value(rhs);
return *this;
}
#endif
inline constexpr MIGRAPHX_HIP_HOST_DEVICE operator float() const
{
return detail::fp8e4m3fnuz_to_fp32_value(data_);
}
fp8e4m3fnuz& MIGRAPHX_HIP_HOST_DEVICE operator=(const fp8e4m3fnuz& rhs) = default;
fp8e4m3fnuz& MIGRAPHX_HIP_HOST_DEVICE operator=(fp8e4m3fnuz&& rhs) = default;
inline bool MIGRAPHX_HIP_HOST_DEVICE isnan() const { return data_ == 0b10000000; }
MIGRAPHX_FP8_UNARY_OP(+=)
MIGRAPHX_FP8_UNARY_OP(-=)
MIGRAPHX_FP8_UNARY_OP(*=)
MIGRAPHX_FP8_UNARY_OP(/=)
friend inline std::ostream& operator<<(std::ostream& out, const fp8e4m3fnuz& value)
{
out << (float)(value);
return out;
}
// what should be the return type ?
MIGRAPHX_FP8_BINARY_OP(+, migraphx::fp8e4m3fnuz)
MIGRAPHX_FP8_BINARY_OP(-, migraphx::fp8e4m3fnuz)
MIGRAPHX_FP8_BINARY_OP(*, migraphx::fp8e4m3fnuz)
MIGRAPHX_FP8_BINARY_OP(/, migraphx::fp8e4m3fnuz)
MIGRAPHX_FP8_BINARY_OP(==, bool)
MIGRAPHX_FP8_BINARY_OP(!=, bool)
MIGRAPHX_FP8_BINARY_OP(>=, bool)
MIGRAPHX_FP8_BINARY_OP(<=, bool)
MIGRAPHX_FP8_BINARY_OP(>, bool)
MIGRAPHX_FP8_BINARY_OP(<, bool)
// implicit conversion should take care of these for the HOST side, half implementation doesn't
// have 'std' implementation MIGRAPHX_FP8_MATH(abs, ::abs) MIGRAPHX_FP8_MATH(acos, ::acos)
// if need to enable these functions, how to put them into std:: namespace ?
// MIGRAPHX_FP8_MATH(acosh, ::acosh)
// MIGRAPHX_FP8_MATH(asin, ::asin)
// MIGRAPHX_FP8_MATH(asinh, ::asinh)
// MIGRAPHX_FP8_MATH(atan, ::atan)
// MIGRAPHX_FP8_MATH(atanh, ::atanh)
// MIGRAPHX_FP8_MATH(ceil, ::ceil)
// MIGRAPHX_FP8_MATH(cos, ::cos)
// MIGRAPHX_FP8_MATH(cosh, ::cosh)
// MIGRAPHX_FP8_MATH(erf, ::erf)
// MIGRAPHX_FP8_MATH(exp, ::exp)
// MIGRAPHX_FP8_MATH(floor, ::floor)
// // MIGRAPHX_FP8_MATH(isnan, ::isnan)
// // MIGRAPHX_FP8_MATH(log, ::log)
// // MIGRAPHX_FP8_MATH(pow, ::pow)
// // MIGRAPHX_FP8_MATH(remainder, ::remainder)
// // MIGRAPHX_FP8_MATH(round, ::round)
// // MIGRAPHX_FP8_MATH(rsqrt, ::rsqrt)
// MIGRAPHX_FP8_MATH(sin, ::sin)
// MIGRAPHX_FP8_MATH(sinh, ::sinh)
// MIGRAPHX_FP8_MATH(sqrt, ::sqrt)
// MIGRAPHX_FP8_MATH(tan, ::tan)
// MIGRAPHX_FP8_MATH(tanh, ::tanh)
// // MIGRAPHX_FP8_MATH(fmod, ::fmod)
};
} // namespace migraphx
namespace std {
template <>
class numeric_limits<migraphx::fp8e4m3fnuz>
{
public:
static constexpr bool is_specialized = true;
static constexpr bool is_signed = true;
static constexpr bool is_integer = false;
static constexpr bool is_exact = false;
static constexpr bool has_infinity = false;
static constexpr bool has_quiet_NaN = true;
static constexpr bool has_signaling_NaN = false;
static constexpr auto has_denorm = true;
static constexpr auto has_denorm_loss = true;
static constexpr auto round_style = numeric_limits<float>::round_style;
static constexpr bool is_iec559 = false;
static constexpr bool is_bounded = true;
static constexpr bool is_modulo = false;
static constexpr int digits = 4;
static constexpr int digits10 = 0;
static constexpr int max_digits10 = 3;
static constexpr int radix = 2;
static constexpr int min_exponent = -5;
static constexpr int min_exponent10 = -1;
static constexpr int max_exponent = 8;
static constexpr int max_exponent10 = 2;
static constexpr auto traps = numeric_limits<float>::traps;
static constexpr auto tinyness_before = false;
static constexpr migraphx::fp8e4m3fnuz min()
{
return migraphx::fp8e4m3fnuz(0x08, migraphx::fp8e4m3fnuz::from_bits());
}
static constexpr migraphx::fp8e4m3fnuz lowest()
{
return migraphx::fp8e4m3fnuz(0xFF, migraphx::fp8e4m3fnuz::from_bits());
}
static constexpr migraphx::fp8e4m3fnuz max()
{
return migraphx::fp8e4m3fnuz(0x7F, migraphx::fp8e4m3fnuz::from_bits());
}
static constexpr migraphx::fp8e4m3fnuz epsilon()
{
return migraphx::fp8e4m3fnuz(0x28, migraphx::fp8e4m3fnuz::from_bits());
}
static constexpr migraphx::fp8e4m3fnuz round_error()
{
return migraphx::fp8e4m3fnuz(0x38, migraphx::fp8e4m3fnuz::from_bits());
}
static constexpr migraphx::fp8e4m3fnuz infinity()
{
// NaN (no infinities)
return migraphx::fp8e4m3fnuz(0x80, migraphx::fp8e4m3fnuz::from_bits());
}
static constexpr migraphx::fp8e4m3fnuz quiet_NaN()
{
return migraphx::fp8e4m3fnuz(0x80, migraphx::fp8e4m3fnuz::from_bits());
}
static constexpr migraphx::fp8e4m3fnuz denorm_min()
{
return migraphx::fp8e4m3fnuz(0x01, migraphx::fp8e4m3fnuz::from_bits());
}
};
template <class T>
struct common_type<migraphx::fp8e4m3fnuz, T> : std::common_type<float, T> // NOLINT
{
};
template <class T>
struct common_type<T, migraphx::fp8e4m3fnuz> : std::common_type<float, T> // NOLINT
{
};
template <>
struct common_type<migraphx::fp8e4m3fnuz, migraphx::fp8e4m3fnuz>
{
using type = float;
};
// template <>
// struct common_type<migraphx::fp8e4m3fnuz, migraphx::half>
// {
// using type = float;
// };
// template <>
// struct common_type<migraphx::half, migraphx::fp8e4m3fnuz>
// {
// using type = float;
// };
} // namespace std
#pragma clang diagnostic pop
#endif
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment