// Generated file (from: l2_normalization_relaxed.mod.py). Do not edit
void CreateModel(Model *model) {
OperandType type0(Type::TENSOR_FLOAT32, {1, 1, 1, 3});
// Phase 1, operands
auto op1 = model->addOperand(&type0);
auto op2 = model->addOperand(&type0);
// Phase 2, operations
model->addOperation(ANEURALNETWORKS_L2_NORMALIZATION, {op1}, {op2});
// Phase 3, inputs and outputs
model->identifyInputsAndOutputs(
{op1},
{op2});
// Phase 4: set relaxed execution
model->relaxComputationFloat32toFloat16(true);
assert(model->isValid());
}
// Generated file (from: transpose.mod.py). Do not edit
void CreateModel(Model *model) {
OperandType type0(Type::TENSOR_FLOAT32, {1, 2, 2, 1});
OperandType type1(Type::TENSOR_INT32, {4});
// Phase 1, operands
auto input = model->addOperand(&type0);
auto perms = model->addOperand(&type1);
auto output = model->addOperand(&type0);
// Phase 2, operations
static int32_t perms_init[] = {0, 2, 1, 3};
model->setOperandValue(perms, perms_init, sizeof(int32_t) * 4);
model->addOperation(ANEURALNETWORKS_TRANSPOSE, {input, perms}, {output});
// Phase 3, inputs and outputs
model->identifyInputsAndOutputs(
{input},
{output});
assert(model->isValid());
}
// Generated file (from: softmax_float_1.mod.py). Do not edit
void CreateModel(Model *model) {
OperandType type1(Type::FLOAT32, {});
OperandType type0(Type::TENSOR_FLOAT32, {1, 4});
// Phase 1, operands
auto input = model->addOperand(&type0);
auto beta = model->addOperand(&type1);
auto output = model->addOperand(&type0);
// Phase 2, operations
static float beta_init[] = {1e-06f};
model->setOperandValue(beta, beta_init, sizeof(float) * 1);
model->addOperation(ANEURALNETWORKS_SOFTMAX, {input, beta}, {output});
// Phase 3, inputs and outputs
model->identifyInputsAndOutputs(
{input},
{output});
assert(model->isValid());
}
// Generated file (from: depth_to_space_float_2.mod.py). Do not edit
void CreateModel(Model *model) {
OperandType type1(Type::INT32, {});
OperandType type0(Type::TENSOR_FLOAT32, {1, 2, 2, 4});
OperandType type2(Type::TENSOR_FLOAT32, {1, 4, 4, 1});
// Phase 1, operands
auto input = model->addOperand(&type0);
auto block_size = model->addOperand(&type1);
auto output = model->addOperand(&type2);
// Phase 2, operations
static int32_t block_size_init[] = {2};
model->setOperandValue(block_size, block_size_init, sizeof(int32_t) * 1);
model->addOperation(ANEURALNETWORKS_DEPTH_TO_SPACE, {input, block_size}, {output});
// Phase 3, inputs and outputs
model->identifyInputsAndOutputs(
{input},
{output});
assert(model->isValid());
}
// Generated file (from: space_to_depth_quant8_2.mod.py). Do not edit
void CreateModel(Model *model) {
OperandType type1(Type::INT32, {});
OperandType type2(Type::TENSOR_QUANT8_ASYMM, {1, 2, 2, 4}, 0.5f, 0);
OperandType type0(Type::TENSOR_QUANT8_ASYMM, {1, 4, 4, 1}, 0.5f, 0);
// Phase 1, operands
auto input = model->addOperand(&type0);
auto radius = model->addOperand(&type1);
auto output = model->addOperand(&type2);
// Phase 2, operations
static int32_t radius_init[] = {2};
model->setOperandValue(radius, radius_init, sizeof(int32_t) * 1);
model->addOperation(ANEURALNETWORKS_SPACE_TO_DEPTH, {input, radius}, {output});
// Phase 3, inputs and outputs
model->identifyInputsAndOutputs(
{input},
{output});
assert(model->isValid());
}
// Generated file (from: mul.mod.py). Do not edit
void CreateModel(Model *model) {
OperandType type1(Type::INT32, {});
OperandType type0(Type::TENSOR_FLOAT32, {1, 2, 2, 1});
// Phase 1, operands
auto op1 = model->addOperand(&type0);
auto op2 = model->addOperand(&type0);
auto act = model->addOperand(&type1);
auto op3 = model->addOperand(&type0);
// Phase 2, operations
static int32_t act_init[] = {0};
model->setOperandValue(act, act_init, sizeof(int32_t) * 1);
model->addOperation(ANEURALNETWORKS_MUL, {op1, op2, act}, {op3});
// Phase 3, inputs and outputs
model->identifyInputsAndOutputs(
{op1, op2},
{op3});
assert(model->isValid());
}
int area(int type)
{
int maxh, maxw;
switch(type) {
/*case 0 : maxw=dane[0].w+dane[1].w+dane[2].w+dane[3].w;
maxh=max(dane[0].h,dane[1].h,dane[2].h,dane[3].h);
return maxw*maxh;*/
case 0 :
return type0();
case 1 :
return type1();
case 2 :
return type2();
case 3 :
return type3();
case 4 :
return type4();
}
}
// Generated file (from: concat_float_2.mod.py). Do not edit
void CreateModel(Model *model) {
OperandType type2(Type::INT32, {});
OperandType type1(Type::TENSOR_FLOAT32, {40, 230});
OperandType type0(Type::TENSOR_FLOAT32, {52, 230});
OperandType type3(Type::TENSOR_FLOAT32, {92, 230});
// Phase 1, operands
auto input1 = model->addOperand(&type0);
auto input2 = model->addOperand(&type1);
auto axis0 = model->addOperand(&type2);
auto output = model->addOperand(&type3);
// Phase 2, operations
static int32_t axis0_init[] = {0};
model->setOperandValue(axis0, axis0_init, sizeof(int32_t) * 1);
model->addOperation(ANEURALNETWORKS_CONCATENATION, {input1, input2, axis0}, {output});
// Phase 3, inputs and outputs
model->identifyInputsAndOutputs(
{input1, input2},
{output});
assert(model->isValid());
}
// Generated file (from: space_to_batch_float_2.mod.py). Do not edit
void CreateModel(Model *model) {
OperandType type0(Type::TENSOR_FLOAT32, {1, 5, 2, 1});
OperandType type3(Type::TENSOR_FLOAT32, {6, 2, 2, 1});
OperandType type2(Type::TENSOR_INT32, {2, 2});
OperandType type1(Type::TENSOR_INT32, {2});
// Phase 1, operands
auto input = model->addOperand(&type0);
auto block_size = model->addOperand(&type1);
auto paddings = model->addOperand(&type2);
auto output = model->addOperand(&type3);
// Phase 2, operations
static int32_t block_size_init[] = {3, 2};
model->setOperandValue(block_size, block_size_init, sizeof(int32_t) * 2);
static int32_t paddings_init[] = {1, 0, 2, 0};
model->setOperandValue(paddings, paddings_init, sizeof(int32_t) * 4);
model->addOperation(ANEURALNETWORKS_SPACE_TO_BATCH_ND, {input, block_size, paddings}, {output});
// Phase 3, inputs and outputs
model->identifyInputsAndOutputs(
{input},
{output});
assert(model->isValid());
}
// Generated file (from: avg_pool_quant8_4.mod.py). Do not edit
void CreateModel(Model *model) {
OperandType type1(Type::INT32, {});
OperandType type0(Type::TENSOR_QUANT8_ASYMM, {1, 3, 3, 1}, 0.5f, 0);
// Phase 1, operands
auto op1 = model->addOperand(&type0);
auto cons1 = model->addOperand(&type1);
auto pad0 = model->addOperand(&type1);
auto relu1_activitation = model->addOperand(&type1);
auto op3 = model->addOperand(&type0);
// Phase 2, operations
static int32_t cons1_init[] = {1};
model->setOperandValue(cons1, cons1_init, sizeof(int32_t) * 1);
static int32_t pad0_init[] = {0};
model->setOperandValue(pad0, pad0_init, sizeof(int32_t) * 1);
static int32_t relu1_activitation_init[] = {2};
model->setOperandValue(relu1_activitation, relu1_activitation_init, sizeof(int32_t) * 1);
model->addOperation(ANEURALNETWORKS_AVERAGE_POOL_2D, {op1, pad0, pad0, pad0, pad0, cons1, cons1, cons1, cons1, relu1_activitation}, {op3});
// Phase 3, inputs and outputs
model->identifyInputsAndOutputs(
{op1},
{op3});
assert(model->isValid());
}
// Generated file (from: conv_3_h3_w2_VALID_relaxed.mod.py). Do not edit
void CreateModel(Model *model) {
OperandType type0(Type::INT32, {});
OperandType type2(Type::TENSOR_FLOAT32, {1, 6, 7, 3});
OperandType type1(Type::TENSOR_FLOAT32, {1, 8, 8, 3});
OperandType type3(Type::TENSOR_FLOAT32, {3, 3, 2, 3});
OperandType type4(Type::TENSOR_FLOAT32, {3});
// Phase 1, operands
auto b4 = model->addOperand(&type0);
auto b5 = model->addOperand(&type0);
auto b6 = model->addOperand(&type0);
auto b7 = model->addOperand(&type0);
auto op2 = model->addOperand(&type1);
auto op3 = model->addOperand(&type2);
auto op0 = model->addOperand(&type3);
auto op1 = model->addOperand(&type4);
// Phase 2, operations
static int32_t b4_init[] = {2};
model->setOperandValue(b4, b4_init, sizeof(int32_t) * 1);
static int32_t b5_init[] = {1};
model->setOperandValue(b5, b5_init, sizeof(int32_t) * 1);
static int32_t b6_init[] = {1};
model->setOperandValue(b6, b6_init, sizeof(int32_t) * 1);
static int32_t b7_init[] = {0};
model->setOperandValue(b7, b7_init, sizeof(int32_t) * 1);
static float op0_init[] = {-0.966213f, -0.579455f, -0.684259f, 0.738216f, 0.184325f, 0.0973683f, -0.176863f, -0.23936f, -0.000233404f, 0.055546f, -0.232658f, -0.316404f, -0.012904f, 0.320705f, -0.326657f, -0.919674f, 0.868081f, -0.824608f, -0.467474f, 0.0278809f, 0.563238f, 0.386045f, -0.270568f, -0.941308f, -0.779227f, -0.261492f, -0.774804f, -0.79665f, 0.22473f, -0.414312f, 0.685897f, -0.327792f, 0.77395f, -0.714578f, -0.972365f, 0.0696099f, -0.82203f, -0.79946f, 0.37289f, -0.917775f, 0.82236f, -0.144706f, -0.167188f, 0.268062f, 0.702641f, -0.412223f, 0.755759f, 0.721547f, -0.43637f, -0.274905f, -0.269165f, 0.16102f, 0.819857f, -0.312008f};
model->setOperandValue(op0, op0_init, sizeof(float) * 54);
static float op1_init[] = {0.0f, 0.0f, 0.0f};
model->setOperandValue(op1, op1_init, sizeof(float) * 3);
model->addOperation(ANEURALNETWORKS_CONV_2D, {op2, op0, op1, b4, b5, b6, b7}, {op3});
// Phase 3, inputs and outputs
model->identifyInputsAndOutputs(
{op2},
{op3});
// Phase 4: set relaxed execution
model->relaxComputationFloat32toFloat16(true);
assert(model->isValid());
}
