v0.3.0

Support for loading models that use the default TensorRT plugins has been added with this release. This new functionality should allow all of the sample Uff applications that ship with the TensorRT library to be ported to Rust!

Other minor changes included with this library are the ability to bind more than one output buffer to a model execution. This allows greater flexibility with creating and executing various model architectures.

Basic Functionality

Still working on rounding out basic functionality but the core functionality has been implemented in this release.

This release also increases the use of ndarray. Specifically around accepting the generic dimension arguments when executing an engine in an execution context.

Version is 0.2.0 is available on crates.io! 🎉
Please download and provide feedback and open issues!