| ONNX_PATH=dinov2_vitl14/1 | |
| MODEL_PATH=model_repository/dinov2_vitl14/1 | |
| model: | |
| git lfs install | |
| git clone git@hf.co:RoundtTble/dinov2_vitl14_onnx | |
| cp -r dinov2_vitl14_onnx/model_repository/dinov2_vitl14/ . | |
| rm -rf dinov2_vitl14_onnx | |
| trt: | |
| mkdir -p $(MODEL_PATH) | |
| docker run --gpus all -it \ | |
| -v $(CURDIR)/$(ONNX_PATH):/workspace/onnx \ | |
| -v $(CURDIR)/$(MODEL_PATH):/workspace/$(MODEL_PATH) \ | |
| --rm nvcr.io/nvidia/pytorch:23.04-py3 \ | |
| /bin/bash -c \ | |
| "trtexec --onnx=onnx/model.onnx --saveEngine=$(MODEL_PATH)/model.plan \ | |
| --minShapes=input:1x3x560x560 \ | |
| --optShapes=input:2x3x560x560 \ | |
| --maxShapes=input:8x3x560x560 \ | |
| --fp16" | |
| triton: | |
| docker run --rm -d \ | |
| --name dinov2_vitl14_triton \ | |
| -p 8000:8000 --gpus "device=0" -p 8001:8001 -p 8002:8002 \ | |
| -v $(PWD)/model_repository:/models \ | |
| nvcr.io/nvidia/tritonserver:23.04-py3 \ | |
| tritonserver --model-repository=/models | |
| perf: | |
| docker run --gpus all --rm -it --net host nvcr.io/nvidia/tritonserver:23.04-py3-sdk perf_analyzer -m dinov2_vitl14 --percentile=95 -i grpc -u 0.0.0.0:8001 --concurrency-range 16:16 --shape input:3,560,560 | |