Entry point to Deep Diamond's CPU engine (currently based on Intel's OneDNNL).
By evaluating this namespace, you're altering uncomplicate.diamond.tensor/*diamond-factory*
var root to the engine produced by [[uncomplicate.diamond.internal.dnnl.factory/dnnl-factory]].
Entry point to Deep Diamond's CPU engine (currently based on Intel's OneDNNL). By evaluating this namespace, you're altering [[uncomplicate.diamond.tensor/*diamond-factory*]] var root to the engine produced by [[uncomplicate.diamond.internal.dnnl.factory/dnnl-factory]].
(find-default-backend)
(load-backend)
(load-backend backend)
(load-bnns)
(load-class classname)
(load-dnnl)
(map-tensor file desc)
(map-tensor file desc flag)
(map-tensor file desc flag offset-bytes)
(map-tensor file shape type format flag)
(map-tensor file shape type format flag offset-bytes)
Creates a new tensor that controls memory block mapped to file
.
Arguments:
shape
, type
, and format
are standard parameters used by Deep Diamonds constructors from the tensor
namespace.flag
s - one of :read-write
, :read
(or :read-only
), :private
(or :copy-on-write
).offset-bytes
- offset in the file where tensor's memory block begins.If you want to see an example of map-tensor
in action, please see how this function is used in
MNIST functional test
and if you need to explore the details, please see the implementation of map channel
in
Neanderthal's internals.
Creates a new tensor that controls memory block mapped to `file`. Arguments: - `shape`, `type`, and `format` are standard parameters used by Deep Diamonds constructors from the `tensor` namespace. - `flag`s - one of `:read-write`, `:read` (or `:read-only`), `:private` (or `:copy-on-write`). - `offset-bytes` - offset in the file where tensor's memory block begins. If you want to see an example of `map-tensor` in action, please see how this function is used in [MNIST functional test](https://github.com/uncomplicate/deep-diamond/blob/master/test/uncomplicate/diamond/functional/mnist/mnist_classification_test.clj) and if you need to explore the details, please see the implementation of `map channel` in [Neanderthal's internals](https://github.com/uncomplicate/neanderthal/blob/master/src/clojure/uncomplicate/neanderthal/internal/cpp/structures.clj).
cljdoc is a website building & hosting documentation for Clojure/Script libraries
× close