1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156
|
#pragma once
#include <torch/csrc/jit/tensorexpr/kernel.h>
namespace torch::jit::tensorexpr {
TORCH_API ExprHandle quantizePerTensorQParamFromArg(ArgValue arg);
TORCH_API double immQScale(const BufHandle& qx);
TORCH_API int64_t immQZero(const BufHandle& qx);
TORCH_API ScalarType immQDType(const BufHandle& qx);
TORCH_API bool isQuantized(const BufHandle& qx);
TORCH_API Tensor computeQuantizePerTensor(
const std::vector<ArgValue>& inputs,
const std::vector<ExprHandle>& outputShape,
const std::vector<ExprHandle>& outputStrides,
const std::optional<ScalarType>& outputType,
at::Device device);
TORCH_API Tensor computeQuantizePerTensorExternalCall(
const std::vector<ArgValue>& inputs,
const std::vector<ExprHandle>& outputShape,
const std::vector<ExprHandle>& outputStrides,
const std::optional<ScalarType>& outputType,
at::Device device);
TORCH_API Tensor computeQuantizedConv1d(
const std::vector<ArgValue>& inputs,
const std::vector<ExprHandle>& outputShape,
const std::vector<ExprHandle>& outputStrides,
const std::optional<ScalarType>& outputType,
at::Device device);
TORCH_API Tensor computeQuantizedConv2dPrepack(
const std::vector<ArgValue>& inputs,
const std::vector<ExprHandle>& outputShape,
const std::vector<ExprHandle>& outputStrides,
const std::optional<ScalarType>& outputType,
at::Device device);
TORCH_API Tensor computeQuantizedConv1d(
const std::vector<ArgValue>& inputs,
const std::vector<ExprHandle>& outputShape,
const std::vector<ExprHandle>& outputStrides,
const std::optional<ScalarType>& outputType,
at::Device device);
TORCH_API Tensor computeQuantizedConv2d(
const std::vector<ArgValue>& inputs,
const std::vector<ExprHandle>& outputShape,
const std::vector<ExprHandle>& outputStrides,
const std::optional<ScalarType>& outputType,
at::Device device);
TORCH_API Tensor computeQuantizedConv2dRelu(
const std::vector<ArgValue>& inputs,
const std::vector<ExprHandle>& outputShape,
const std::vector<ExprHandle>& outputStrides,
const std::optional<ScalarType>& outputType,
at::Device device);
TORCH_API Tensor computeQuantizedLinear(
const std::vector<ArgValue>& inputs,
const std::vector<ExprHandle>& outputShape,
const std::vector<ExprHandle>& outputStrides,
const std::optional<ScalarType>& outputType,
at::Device device);
TORCH_API Tensor computeQuantizedLinearRelu(
const std::vector<ArgValue>& inputs,
const std::vector<ExprHandle>& outputShape,
const std::vector<ExprHandle>& outputStrides,
const std::optional<ScalarType>& outputType,
at::Device device);
TORCH_API Tensor computeQuantizedAdd(
const std::vector<ArgValue>& inputs,
const std::vector<ExprHandle>& outputShape,
const std::vector<ExprHandle>& outputStrides,
const std::optional<ScalarType>& outputType,
at::Device device);
Tensor computeQuantizedAddExternalCall(
const std::vector<ArgValue>& inputs,
const std::vector<ExprHandle>& outputShape,
const std::vector<ExprHandle>& outputStrides,
const std::optional<ScalarType>& outputType,
at::Device device);
TORCH_API Tensor computeQuantizedMul(
const std::vector<ArgValue>& inputs,
const std::vector<ExprHandle>& outputShape,
const std::vector<ExprHandle>& outputStrides,
const std::optional<ScalarType>& outputType,
at::Device device);
TORCH_API Tensor computeQuantizedMulScalar(
const std::vector<ArgValue>& inputs,
const std::vector<ExprHandle>& outputShape,
const std::vector<ExprHandle>& outputStrides,
const std::optional<ScalarType>& outputType,
at::Device device);
TORCH_API Tensor computeQuantizedCat(
const std::vector<ArgValue>& inputs,
const std::vector<ExprHandle>& outputShape,
const std::vector<ExprHandle>& outputStrides,
const std::optional<ScalarType>& outputType,
at::Device device);
TORCH_API Tensor computeQuantizedRelu(
const std::vector<ArgValue>& inputs,
const std::vector<ExprHandle>& outputShape,
const std::vector<ExprHandle>& outputStrides,
const std::optional<ScalarType>& outputType,
at::Device device);
TORCH_API Tensor computeDequantize(
const std::vector<ArgValue>& inputs,
const std::vector<ExprHandle>& outputShape,
const std::vector<ExprHandle>& outputStrides,
const std::optional<ScalarType>& outputType,
at::Device device);
TORCH_API Tensor computeDequantizeExternalCall(
const std::vector<ArgValue>& inputs,
const std::vector<ExprHandle>& outputShape,
const std::vector<ExprHandle>& outputStrides,
const std::optional<ScalarType>& outputType,
at::Device device);
TORCH_API Tensor computeUpsampleNearest2d(
const std::vector<ArgValue>& inputs,
const std::vector<ExprHandle>& outputShape,
const std::vector<ExprHandle>& outputStrides,
const std::optional<ScalarType>& outputType,
at::Device device);
TORCH_API Tensor computeUpsampleNearest2dExternalCall(
const std::vector<ArgValue>& inputs,
const std::vector<ExprHandle>& outputShape,
const std::vector<ExprHandle>& outputStrides,
const std::optional<ScalarType>& outputType,
at::Device device);
TORCH_API Tensor computeQuantizedSigmoidExternalCall(
const std::vector<ArgValue>& inputs,
const std::vector<ExprHandle>& outputShape,
const std::vector<ExprHandle>& outputStrides,
const std::optional<ScalarType>& outputType,
at::Device);
} // namespace torch::jit::tensorexpr
|