nntrainer
nntrainer copied to clipboard
nnstreamer
NNtrainer is Software Framework for Training Neural Network Models on Devices.
- 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....
change context of train log file and point of file read/write
We need to provide the getter api for layers, models and optimizers. Once user set the properties, there is no way to check except model_summary.
NNtrainer generates the binary file for the inference. It depends on the scenario but there is request to use saved binary file again to fine tuning the model.
Later, `-Wno-unused-variable` can be removed once the underlying issue is resolved, or added to the specific files causing this issue. _Originally posted by @kparichay in https://github.com/nnstreamer/nntrainer/pull/1858#pullrequestreview-914989329_ After addressing all -Wno-unused-variables...
The public API requires the optimizer/dataset to be set before ml_train_model_compile. However, these restrictions can be relaxed and both can be later right until the training is being done: 1....
Adding a layer into a model created with INI requires creating two connections for the new layer - with the previous layer and with the next layer. In order to...
The current model save/load file format stores the tensors in a linear fashion. This has a couple of issues: 1. re-ordering the model layers results in failure 2. adding a...
Current inference API requires the caller to provide individual inputs to the inference call. This requires handling the inputs, outputs, and file reading on the user's end. The proposal is...
Adam variables are loaded with `ml::train::model::load()`, which allows loading the model weights. However, model weights can be loaded with a call to `ml::train::model::load()` but adam weights can only be loaded...
Metadata
Owner
Metadata
NNtrainer is Software Framework for Training Neural Network Models on Devices.