The doc introduces how to convert models into TNN.
step 1, convert your model to ONNX, run:
python3 tools/convert2onnx.py <config-file> --input-img <img-dir> --shape 512 512 --checkpoint <model-ckpt>
step 2, clone the TNN:
git clone https://github.com/Tencent/TNN.git
step 3, covert model to TNN following convert2tnn.
- install the dependencies of convert2tnn
- compile the tools.
cd <path-to-tnn>/tools/onnx2tnn/onnx-converter
./build.sh
then run:
python3 converter.py onnx2tnn <onnx-model-path> -optimize -v=v3.0 -o <output-file-name>
step 4, compile for Android following tnn-compile.
- cmake（version 3.6 or higher）
- NDK configuration
then run:
./build_android.sh
step 5, models benchmark.
push all benchmark models to android device_dir:
/data/local/tmp/benchmark-model
then run:
./benchmark_models.sh -32 -t cpu -bs -f
you will get all model benchmark cost time info.