From 0b474a55a2a1dc123c721db3e934af72e585a2ad Mon Sep 17 00:00:00 2001
From: thibault allenet <thibault.allenet@cea.fr>
Date: Wed, 6 Nov 2024 12:56:52 +0000
Subject: [PATCH] Add type aidge::int4 with tensor implementation int8_t
---
include/aidge/backend/cpu/data/TensorImpl.hpp | 1 +
include/aidge/data/Data.hpp | 11 +++++++++++
python_binding/data/pybind_Tensor.cpp | 5 +++++
src/data/Tensor.cpp | 2 ++
4 files changed, 19 insertions(+)
diff --git a/include/aidge/backend/cpu/data/TensorImpl.hpp b/include/aidge/backend/cpu/data/TensorImpl.hpp
index 9390fe586..5f4e25772 100644
--- a/include/aidge/backend/cpu/data/TensorImpl.hpp
+++ b/include/aidge/backend/cpu/data/TensorImpl.hpp
@@ -126,6 +126,7 @@ REGISTRAR(Tensor, {"cpu", DataType::Int64}, Aidge::TensorImpl_cpu<int64_t>::crea
REGISTRAR(Tensor, {"cpu", DataType::Int32}, Aidge::TensorImpl_cpu<int32_t>::create);
REGISTRAR(Tensor, {"cpu", DataType::Int16}, Aidge::TensorImpl_cpu<int16_t>::create);
REGISTRAR(Tensor, {"cpu", DataType::Int8}, Aidge::TensorImpl_cpu<int8_t>::create);
+REGISTRAR(Tensor, {"cpu", DataType::Int4}, Aidge::TensorImpl_cpu<int8_t>::create);
REGISTRAR(Tensor, {"cpu", DataType::UInt64}, Aidge::TensorImpl_cpu<uint64_t>::create);
REGISTRAR(Tensor, {"cpu", DataType::UInt32}, Aidge::TensorImpl_cpu<uint32_t>::create);
REGISTRAR(Tensor, {"cpu", DataType::UInt16}, Aidge::TensorImpl_cpu<uint16_t>::create);
diff --git a/include/aidge/data/Data.hpp b/include/aidge/data/Data.hpp
index 23221e653..d4d206409 100644
--- a/include/aidge/data/Data.hpp
+++ b/include/aidge/data/Data.hpp
@@ -101,10 +101,17 @@ private:
}
namespace {
+
+// Define a distinct type alias for Int4
+struct Int4Type {
+ std::int8_t value;
+};
+
template <typename T> struct NativeType { static const Aidge::DataType type; };
template <> const Aidge::DataType NativeType<double>::type = Aidge::DataType::Float64;
template <> const Aidge::DataType NativeType<float>::type = Aidge::DataType::Float32;
template <> const Aidge::DataType NativeType<half_float::half>::type = Aidge::DataType::Float16;
+template <> const Aidge::DataType NativeType<Int4Type>::type = Aidge::DataType::Int4;
template <> const Aidge::DataType NativeType<std::int8_t>::type = Aidge::DataType::Int8;
template <> const Aidge::DataType NativeType<std::int16_t>::type = Aidge::DataType::Int16;
template <> const Aidge::DataType NativeType<std::int32_t>::type = Aidge::DataType::Int32;
@@ -131,6 +138,7 @@ template <Aidge::DataType D> struct cpptype {
template <> struct cpptype<Aidge::DataType::Float16> { using type = half_float::half; };
template <> struct cpptype<Aidge::DataType::Float32> { using type = float; };
template <> struct cpptype<Aidge::DataType::Float64> { using type = double; };
+template <> struct cpptype<Aidge::DataType::Int4> { using type = Int4Type; };
template <> struct cpptype<Aidge::DataType::Int8> { using type = std::int8_t; };
template <> struct cpptype<Aidge::DataType::Int16> { using type = std::int16_t; };
template <> struct cpptype<Aidge::DataType::Int32> { using type = std::int32_t; };
@@ -141,6 +149,9 @@ template <> struct cpptype<Aidge::DataType::UInt32> { using type = std::uint32_t
template <> struct cpptype<Aidge::DataType::UInt64> { using type = std::uint64_t; };
template <Aidge::DataType D> using cpptype_t = typename cpptype<D>::type;
+
+
+
}
diff --git a/python_binding/data/pybind_Tensor.cpp b/python_binding/data/pybind_Tensor.cpp
index fe606cfb5..7e910be57 100644
--- a/python_binding/data/pybind_Tensor.cpp
+++ b/python_binding/data/pybind_Tensor.cpp
@@ -226,6 +226,8 @@ static T castToNativeType(const py::object val_obj) {
DataType dtype;
getConservativeNativeVal(val_obj, &val, &dtype);
switch (dtype) {
+ case DataType::Int4:
+ return (T)val.i8;
case DataType::Int8:
return (T)val.i8;
case DataType::Int16:
@@ -500,6 +502,9 @@ void init_Tensor(py::module& m){
case DataType::Float32:
dataFormatDescriptor = py::format_descriptor<float>::format();
break;;
+ case DataType::Int4:
+ dataFormatDescriptor = py::format_descriptor<std::int8_t>::format();
+ break;
case DataType::Int8:
dataFormatDescriptor = py::format_descriptor<std::int8_t>::format();
break;
diff --git a/src/data/Tensor.cpp b/src/data/Tensor.cpp
index 6f60d2f15..2ba19a749 100644
--- a/src/data/Tensor.cpp
+++ b/src/data/Tensor.cpp
@@ -249,6 +249,8 @@ std::string Aidge::Tensor::toString() const {
return std::to_string(static_cast<float*>(ptr)[idx]);
case DataType::Float16:
return std::to_string(static_cast<half_float::half*>(ptr)[idx]);
+ case DataType::Int4:
+ return std::to_string(static_cast<int8_t*>(ptr)[idx]);
case DataType::Int8:
return std::to_string(static_cast<int8_t*>(ptr)[idx]);
case DataType::Int16:
--
GitLab