Module ops
Expand description
Operations on tensors module.
Modules§
- Module with convolution operations.
- Module with pooling operations.
Structs§
- Gradient computed during the backward pass for each tensor used by conv2d.
- Gradient computed during the backward pass for each tensor used by conv3d.
- Convolution options.
- Transposed convolution options.
- Gradient computed during the backward pass for each tensor used by deform_conv2d.
- Convolution options.
- Gradient computed during the backward pass for each tensor used by interpolate.
- Interpolation options.
- Gradient computed during the backward pass for each tensor used by max_pool1d.
- Results from max_pool1d.
- Gradient computed during the backward pass for each tensor used by max_pool2d.
- Results from max_pool2d.
- Contains all tensor primitives that are going to be read.
- Contains all data related to a transaction.
- Unfold operation options.
Enums§
- Algorithm used for upsampling.
Traits§
- Activation function operations.
- Bool Tensor API for basic operations, see tensor for documentation on each function.
- Operations on float tensors.
- Int Tensor API for basic and numeric operations, see tensor for documentation on each function.
- Module operations trait.
- Quantized Tensor API for basic operations, see tensor for documentation on each function.
- Operations that are sync by nature and that can be batch together in transactions to improve compute utilization with efficient laziness.
Functions§
- Computes the output shape for binary operations with broadcasting support.
Type Aliases§
- Boolean tensor primitive type used by the backend.
- Device type used by the backend.
- Float element type used by backend.
- Float tensor primitive type used by the backend.
- Integer element type used by backend.
- Integer tensor primitive type used by the backend.
- Quantized tensor primitive type used by the backend.