ONNX-TensorRT: TensorRT backend for ONNX
For more details, see the 8.6 GA release notes for new features added in TensorRT 8.6.
GroupNormalization
, LayerNormalization
,IsInf
operationsArgmin
, Argmax
, and TopK
ReverseSequence
operators with dynamic shapesTopK
operators with dynamic K valuesTrilu
OnnxParserFlag
enum and setFlag
interfaces to the ONNX parser to modify the default parsing behaviorFor more details, see the 8.6 EA release notes for new features added in TensorRT 8.6
GroupNormalization
, LayerNormalization
, IsInf
operationsArgmin
, Argmax
, and TopK
ReverseSequence
operators with dynamic shapesTopK
operators with dynamic K
valuesOnnxParserFlag
enum and setFlag
interfaces to the ONNX parser to modify the default parsing behaviorCastLayer
over the pervious IdentityLayer
.For more details, see the 8.5 GA release notes for new features added in TensorRT 8.5
RandomNormal
, RandomUniform
, MeanVarianceNormalization
, RoiAlign
, Mod
, Trilu
, GridSample
and NonZero
operationsNonMaxSuppression
operatorUINT8
I/O typesIf
subgraphs with shared initializer namesLoop
subgraphs with INT_MAX
trip countsonnx2trt
binary. See the README.md for alternative binaries to run ONNX model with TensorRT.For more details, see the 8.4 GA release notes for new features added in TensorRT 8.4
Shrink
Xor
ArgMax
and ArgMin
Range
for floating point typesCelu
importer functionGEMM
importer functionGather
and GatherElements
implementations to natively support negative indicesPad
layer to support ND padding, along with edge
and reflect
padding mode supportIf
layer with general performance improvements.half_pixel
, pytorch_half_pixel
, tf_half_pixel_for_nn
, asymmetric
, and align_corners
nearest
, linear
floor
, ceil
, round_prefer_floor
, round_prefer_ceil
NonMaxSuppression
RefitMap
API to core TensorRT.For more information on the TensorRT 8.0 release, see the release notes.
DOUBLE
-typed inputs and weights through downcast to float #674