Searched refs:tensor (Results 1 – 12 of 12) sorted by relevance
/hardware/interfaces/neuralnetworks/1.1/ |
D | types.hal | 39 * Supported tensor {@link OperandType}: 43 * Supported tensor rank: 4, with "NHWC" (i.e., Num_samples, Height, Width, 47 * * 0: An n-D tensor, specifying the tensor to be reshaped 49 * sizes for each spatial dimension of the input tensor. All values 53 * * 0: A tensor of the same {@link OperandType} as input0. 54 * For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor, 63 * dimensions. The output is the result of dividing the first input tensor 79 * Supported tensor {@link OperandType}: 82 * Supported tensor rank: up to 4 85 * * 0: An n-D tensor, specifying the first input. [all …]
|
/hardware/interfaces/neuralnetworks/1.2/ |
D | types.hal | 49 * A tensor of 16 bit signed integers that represent real numbers. 51 * Attached to this tensor is a number representing real value scale that is 59 * A tensor of IEEE 754 16 bit floating point values. 63 * A tensor of 8 bit boolean values. 74 * A tensor of 8 bit signed integers that represent real numbers. 76 * This tensor is associated with additional fields that can 84 * The channel dimension of this tensor must not be unknown (dimensions[channelDim] != 0). 93 * A tensor of 16 bit unsigned integers that represent real numbers. 95 * Attached to this tensor are two numbers that can be used to convert the 105 * A tensor of 8 bit signed integers that represent real numbers. [all …]
|
D | types.t | 207 * A tensor operand with all dimensions specified has "fully 209 * dimensions are known at model construction time), a tensor 214 * If a tensor operand's dimensions are not fully specified, the 219 * In the following situations, a tensor operand's dimensions must 234 * A tensor operand with some number of unspecified dimensions is 237 * A tensor operand with unspecified rank is represented by providing 712 * Indicates whether the extension operand type represents a tensor or 719 * tensor).
|
D | IPreparedModel.hal | 57 * If the prepared model was prepared from a model wherein all tensor 109 * If the prepared model was prepared from a model wherein all tensor 152 * If the prepared model was prepared from a model wherein all tensor
|
D | IDevice.hal | 133 * and transformed tensor buffers. Any modification to the data cache should
|
/hardware/interfaces/neuralnetworks/1.3/ |
D | types.hal | 33 * A tensor of 8 bit signed integers that represent real numbers. 35 * Attached to this tensor are two numbers that can be used to convert the 63 * A tensor of OEM specific values. 114 * Since HAL version 1.2, generic zero-sized input tensor is supported. Zero 118 * Supported tensor {@link OperandType}: 125 * Supported tensor rank: up to 4 128 * * 0: A tensor. 129 * * 1: A tensor of the same {@link OperandType}, and compatible dimensions 132 * {@link OperandType::TENSOR_QUANT8_ASYMM_SIGNED} tensor, 137 * For a {@link OperandType::TENSOR_INT32} tensor, [all …]
|
D | types.t | 257 * A tensor operand with all dimensions specified has "fully 259 * dimensions are known at model construction time), a tensor 264 * If a tensor operand's dimensions are not fully specified, the 271 * In the following situations, a tensor operand's dimensions must 286 * A tensor operand with some number of unspecified dimensions is 289 * A tensor operand with unspecified rank is represented by providing
|
D | IPreparedModel.hal | 62 * If the prepared model was prepared from a model wherein all tensor 140 * If the prepared model was prepared from a model wherein all tensor
|
/hardware/interfaces/neuralnetworks/1.0/ |
D | types.hal | 24 * Types prefaced with TENSOR_* must be used for tensor data (i.e., tensors 40 /** A tensor of 32 bit floating point values. */ 42 /** A tensor of 32 bit integer values. */ 45 * A tensor of 8 bit unsigned integers that represent real numbers. 47 * Attached to this tensor are two numbers that can be used to convert the 69 * A tensor of OEM specific values. 101 * Supported tensor {@link OperandType}: 105 * Supported tensor rank: up to 4 108 * * 0: A tensor. 109 * * 1: A tensor of the same {@link OperandType}, and compatible dimensions [all …]
|
D | types.t | 37 * A tensor of OEM specific values. 192 * For a tensor operand, dimensions.size() must be at least 1; 195 * A tensor operand with all dimensions specified has "fully 197 * dimensions are known at model construction time), a tensor 202 * If a tensor operand's dimensions are not fully specified, the 207 * In the following situations, a tensor operand's dimensions must 222 * A tensor operand with some number of unspecified dimensions is
|
D | IPreparedModel.hal | 52 * If the prepared model was prepared from a model wherein all tensor
|
/hardware/qcom/neuralnetworks/hvxservice/1.0/ |
D | HexagonModel.cpp | 158 hexagon_nn_input& tensor = mOperands[operand].hexagon_input; in getTensor() local 159 if (tensor == hexagon_nn_input{}) { in getTensor() 160 tensor = addOperand(operand); in getTensor() 162 return tensor; in getTensor()
|