Searched refs:num_units (Results 1 – 4 of 4) sorted by relevance
| /hardware/interfaces/neuralnetworks/1.0/ |
| D | types.hal | 554 * [num_units, input_size], where "num_units" corresponds to the number
556 * * 2: A 1-D tensor, of shape [num_units], specifying the bias. For input
567 * * 0: The output tensor, of shape [batch_size, num_units]. For
933 * A 2-D tensor of shape [num_units, input_size], where “num_units”
936 * A 2-D tensor of shape [num_units, input_size].
938 * A 2-D tensor of shape [num_units, input_size].
940 * A 2-D tensor of shape [num_units, input_size].
942 * A 2-D tensor of shape [num_units, output_size], where “output_size”
943 * corresponds to either the number of cell units (i.e., “num_units”),
946 * A 2-D tensor of shape [num_units, output_size].
[all …]
|
| /hardware/qcom/neuralnetworks/hvxservice/1.0/ |
| D | HexagonModel.cpp | 236 uint32_t num_units = dims[0] * dims[1] * dims[2]; in createFullyConnectedWeightTensor() local
240 num_units, input_size, reinterpret_cast<const float*>(operandInfo.buffer)); in createFullyConnectedWeightTensor()
241 return createTensorInternal(1, 1, input_size, num_units, in createFullyConnectedWeightTensor()
246 num_units, input_size, reinterpret_cast<const uint8_t*>(operandInfo.buffer)); in createFullyConnectedWeightTensor()
247 return createTensorInternal(1, 1, input_size, num_units, in createFullyConnectedWeightTensor()
|
| /hardware/interfaces/neuralnetworks/1.2/ |
| D | types.hal | 772 * [num_units, input_size], where "num_units" corresponds to the number
774 * * 2: A 1-D tensor, of shape [num_units], specifying the bias. For input
785 * * 0: The output tensor, of shape [batch_size, num_units]. Before HAL version 1.2, for
1218 * A 2-D tensor of shape [num_units, input_size], where “num_units”
1221 * A 2-D tensor of shape [num_units, input_size].
1223 * A 2-D tensor of shape [num_units, input_size].
1225 * A 2-D tensor of shape [num_units, input_size].
1227 * A 2-D tensor of shape [num_units, output_size], where “output_size”
1228 * corresponds to either the number of cell units (i.e., “num_units”),
1231 * A 2-D tensor of shape [num_units, output_size].
[all …]
|
| /hardware/interfaces/neuralnetworks/1.3/ |
| D | types.hal | 758 * [num_units, input_size], where "num_units" corresponds to the number
760 * * 2: A 1-D tensor, of shape [num_units], specifying the bias. For input
772 * * 0: The output tensor, of shape [batch_size, num_units]. Before HAL version 1.2, for
1215 * A 2-D tensor of shape [num_units, input_size], where “num_units”
1218 * A 2-D tensor of shape [num_units, input_size].
1220 * A 2-D tensor of shape [num_units, input_size].
1222 * A 2-D tensor of shape [num_units, input_size].
1224 * A 2-D tensor of shape [num_units, output_size], where “output_size”
1225 * corresponds to either the number of cell units (i.e., “num_units”),
1228 * A 2-D tensor of shape [num_units, output_size].
[all …]
|