The plugins belong in the nvinfer1::plugin namespace.
However some plugins were not only polluting the global namespace, but
were exposing the same function twice! Once in the global namespace and
once in the nvinfer1::plugin namespace. This fixes this by pushing
everything into the nvinfer1::plugin namespace.
Signed-off-by: Shuyue Lan <shuyuel@nvidia.com>
Signed-off-by: Rajeev Rao <rajeevrao@nvidia.com>
For version forward compatibility, we want to generate a shared library which contains vfc plugins. Currently we only have ROIAlign plugin in the list.
Signed-off-by: Shuyue Lan <shuyuel@nvidia.com>
Signed-off-by: Rajeev Rao <rajeevrao@nvidia.com>
Update tools including onnx-graphsurgeon(v0.3.25), Polygraphy(v0.43.1), tensorflow-quantization(v0.2.0)
Update onnx-tensorrt(8.5GA)
Mass merge sample, plugin, demo, python changes for TRT-8.5 GA release
Update demoBERT benchmark data for TensorRT 8.5.1
Signed-off-by: Simeng Liu <simengl@nvidia.com>
Signed-off-by: Simeng Liu <simengl@nvidia.com>
This graceful failure is needed to enable tests under `trt_mod_test`.
`PLUGIN_ASSERT` is used in many places inside plugin methods, most of
which are `noexcept` and do not contain try-catch blocks. So changing
`PLUGIN_ASSERT` would require a major refactor which is beyond
rel-8.4. Instead, this MR introduces `PLUGIN_VALIDATE` as a temporary
replacement for `PLUGIN_ASSERT` in places where it is required to enable
valid/invalid attribute tests (mostly `createPlugin` and plugin constructors).
i.e. `PLUGIN_VALIDATE` implements the expected behavior of `PLUGIN_ASSERT`
(throws) while `PLUGIN_ASSERT` still aborts.
Signed-off-by: samurdhi karunaratne <skarunaratne@nvidia.com>
Signed-off-by: Rajeev Rao <rajeevrao@nvidia.com>
Moves `BasePlugin` and `BaseCreator` in `plugin/common/plugin.h` out of
`nvinfer1::plugin` into `nvinfer1::pluginInternal`, as currently,
everything inside `nvinfer1::plugin` is being exported and those two
are meant to be internal helper classes.
Signed-off-by: samurdhi karunaratne <skarunaratne@nvidia.com>
Signed-off-by: Rajeev Rao <rajeevrao@nvidia.com>
Kevin Chen