Expand description
This crate is dynamic graph based tensor library
Modules§
- alloc
- module for memory allocation
- backend
- module for backend
- buitin_
templates - module for buitin templates
- common
- module for common utils like shape and strides
- error
- module for error handling
- iter
- module for tensor iterator
- ops
- this module contains all the operators for the Tensor
- re_
exports - reexport serde
- save_
load - module for save and load
- types
- type related module
- utils
- module for utils, like set_num_threads, set_seed, etc.
Macros§
- conv2d_
micro_ kernel - conv2d micro kernel
- conv2d_
micro_ kernel_ with_ padding - conv2d micro kernel with padding
- conv2d_
mixed_ precision_ micro_ kernel - conv2d mixed precision micro kernel
- conv2d_
mixed_ precision_ micro_ kernel_ with_ padding - conv2d mixed precision micro kernel
- define_
matmul_ micro_ kernel - Define Matmul Micro Kernel function
- define_
mixed_ precision_ matmul_ micro_ kernel - Define Matmul Micro Kernel function
- define_
mixed_ precision_ post_ op_ matmul_ micro_ kernel - Define Matmul Micro Kernel function with post operation
- define_
post_ op_ matmul_ micro_ kernel - Define Matmul Micro Kernel function with post operation
- slice
- slice operation for tensor slicing uses the same syntax as numpy
Structs§
- Tensor
Tensoris alias of N-dimensional array.