machine-learning/triton/amazon-sagamaker-triton-resnet-50/set_up/workspace/generate_model_trt.sh
#!/usr/bin/env bash
set -e
python export_onnx.py
# Generate a model plan that will be used to host SageMaker Endpoint
trtexec \
--onnx=model.onnx \
--saveEngine=model.plan \
--minShapes=input:1x3x224x224 \
--optShapes=input:128x3x224x224 \
--maxShapes=input:128x3x224x224 \
--explicitBatch \
--fp16 \
--verbose \
| tee conversion.txt