Open Hukongtao opened 3 months ago
Did the code examples provided by the official documentation also have errors?
Thanks for catching this. The correct usage is as follows:
trt_gm = torch_tensorrt.dynamo.compile(exp_program, inputs)
# Run inference
trt_gm(*inputs)
inputs is already a list and [inputs]
was being passed which is incorrect. Raised a PR here : https://github.com/pytorch/TensorRT/pull/3036
Bug Description
To Reproduce
The code comes from the official documentation: https://pytorch.org/TensorRT/user_guide/dynamic_shapes.html#custom-dynamic-shape-constraints
Expected behavior
run successfully
Environment
conda
,pip
,libtorch
, source):Additional context