-if [ -z "$TARGET_SOCS" ]; then TARGET_SOCS=random; fi
->
for CONF_FILE in mace-models/mobilenet-v1/mobilenet-v1-quantize-friendly.yml mace-models/mobilenet-v1/mobilenet-v1-quantize-retrain-for-check-only.yml mace-models/mobilenet-v1/mobilenet-v1-quantize-retrain-dsp.yml;
@@ -72,7 +72,7 @@ If model's output is suspected to be incorrect, it might be useful to debug your
...
@@ -72,7 +72,7 @@ If model's output is suspected to be incorrect, it might be useful to debug your
or use binary search method until suspicious layer is found.
or use binary search method until suspicious layer is found.
You can also specify `--layers` after `--validate` to validate all or some of the layers of the model(excluding some layers changed by MACE, e.g., BatchToSpaceND),
You can also specify `--layers` after `--validate` to validate all or some of the layers of the model(excluding some layers changed by MACE, e.g., BatchToSpaceND),
it only supports TensorFlow now. You can find validation results in `builds/your_model/model/runtime_in_yaml/log.csv`.
it only supports TensorFlow now. You can find validation results in `build/your_model/model/runtime_in_yaml/log.csv`.
For quantized model, if you want to check one layer, you can add `check_tensors` and `check_shapes` like in the yaml above. You can only specify
For quantized model, if you want to check one layer, you can add `check_tensors` and `check_shapes` like in the yaml above. You can only specify
MACE op's output.
MACE op's output.
...
@@ -113,7 +113,7 @@ Debug model conversion
...
@@ -113,7 +113,7 @@ Debug model conversion
After model is converted to MACE model, a literal model graph is generated in directory `mace/codegen/models/your_model`.
After model is converted to MACE model, a literal model graph is generated in directory `mace/codegen/models/your_model`.
You can refer to it when debugging model conversion.
You can refer to it when debugging model conversion.
MACE also provides model visualization HTML generated in `builds` directory, generated after converting model.
MACE also provides model visualization HTML generated in `build` directory, generated after converting model.
This command will download or load your pre-trained model and convert it to a MACE model proto file and weights data file.
This command will download or load your pre-trained model and convert it to a MACE model proto file and weights data file.
The generated model files will be stored in ``builds/${library_name}/model`` folder.
The generated model files will be stored in ``build/${library_name}/model`` folder.
.. warning::
.. warning::
...
@@ -188,10 +188,10 @@ Or use bazel to build MACE source code into a library.
...
@@ -188,10 +188,10 @@ Or use bazel to build MACE source code into a library.
cd path/to/mace
cd path/to/mace
# Build library
# Build library
# output lib path: builds/lib
# output lib path: build/lib
bash tools/build-standalone-lib.sh
bash tools/bazel-build-standalone-lib.sh
The above command will generate dynamic library ``builds/lib/${ABI}/${DEVICES}/libmace.so`` and static library ``builds/lib/${ABI}/${DEVICES}/libmace.a``.
The above command will generate dynamic library ``build/lib/${ABI}/${DEVICES}/libmace.so`` and static library ``build/lib/${ABI}/${DEVICES}/libmace.a``.