Po-Wei (Vincent) comments

Results 70 comments of


                                            Po-Wei (Vincent)

Error when creating TRT engine with INT8 calibration

Closing due to inactive. Please feel free to reopen!

NO Tensorrt 9.2

Closing based on https://github.com/NVIDIA/TensorRT/issues/4200#issuecomment-2415790867. Feel free to reopen 👍

Engine build failure "cuda misaligned address" of TensorRT 10.1 when running fp16 group normalization with particular value of `num_groups`

Closing due to inactive. Please feel free to reopen!

How can I transfer t5 decoder_model_merged.onnx to tensorrt

Closing this issue for now but feel free to reopen if you see further issues!

set fp32 for MatMul failure of TensorRT 8.6.1.1 when converting onnx to trt with version 8.6.1.1

Closing due to inactive. Please feel free to reopen!

INT8 Quantization of a custom model failed

Closing this issue for now but feel free to reopen the issue if you still see the error for trt >= 10.0 👍

The KL divergence calculation is very slow and is not optimized for acceleration

Closing this issue for now for being stale but feel free to reopen if you still see the problem!

"index with mask" failure of TensorRT 8.6 when running trtexec on GPU A4500

Hello there's a sample for data-dependent shape plugins [here](https://github.com/NVIDIA/TensorRT/tree/release/10.8/samples/sampleNonZeroPlugin). You can also override [IOutputAllocator](https://docs.nvidia.com/deeplearning/tensorrt/api/c_api/classnvinfer1_1_1v__1__0_1_1_i_output_allocator.html) for this purpose. Closing for now but feel free to reopen if you still see the...

AttributeError: 'tensorrt_bindings.tensorrt.ICudaEngine' object has no attribute 'get_binding_index'

Closing due to inactive. Please feel free to reopen!

The onnx parser failed to parse a valid model: Slice (importSlice): INVALID_NODE: Assertion failed: (starts.size() == axes.size())

> [05/29/2025-10:49:12] [TRT] [E] In node 5 with name: and operator: Slice (importSlice): INVALID_NODE: Assertion failed: (starts.size() == axes.size()): The shape of input starts misaligns with the shape of input...