Jijoong Moon
Jijoong Moon
When I use the PR Report Handler in "Standalone Continuous Integration Server" Web Interface, there is shell command exposed as below. ```bash ID and Pasword is correct. Running the PR...
This patch includes implementation of sgemv function with ARM NEON. Later, more arm neon optimized function will be added **Self evaluation:** 1. Build test: [X]Passed [ ]Failed [ ]Skipped 2....
We need to verify once again when we using ini format model. 1. Create Model with C++ API 2. call ``` c++ model->save('exported.ini', ml::train::ModelFormat::MODEL_FORMAT_INI)``` 3. run model with ``` c++...
We need to support `addWithReferenceLayers` API in ini Configuration file format. ```c++ /** * @brief add refering to reference layers. * @note This method does add the provided layers itself...
Currently, NNtrainer uses the ` addWithReferenceLayers(...)` to unroll the subGraph. The maximum time of unroll is defined using `unroll_for` property and every memory space including tensors is allocated accordingly. However,...
There are python scripts to generate the golden data for the test. ``` bash $ pwd ${NNTRAINER_HOME}/test/input_gen $ ls *.py compute_grad.py genModelsMultiout_v2.py recorder.py transLayer_v2.py embedding.py genModelsRecurrent_v2.py recorder_v2.py zoneout.py genInput.py genModelTests.py...
This might be helpful. (https://numpy.org/doc/stable/reference/generated/numpy.matmul.html) According to the description
Currently, NNtrainer only supports the file type for the dataset in ini file. However, we also have the dataset API for the callback and It needs to be supported in...
- Develop Face Lock Tizen Application
- NNTrainer needs to export the trained model to TfLite. - Currently, we have PoC Code for Fullyconnected layer, but we need to developer tflite interpreter to support more layers....