tflite-micro
tflite-micro copied to clipboard
tensorflow
Infrastructure to enable deployment of ML models to low-power resource-constrained embedded targets (including microcontrollers and digital signal processors).
The [MNIST LSTM example](https://github.com/tensorflow/tflite-micro/tree/main/tensorflow/lite/micro/examples/mnist_lstm) does not compile a correct TFLite model with the UnidirectionalSequenceLSTM operator for TFLite Micro. Instead it generates a large graph with many operators. Versions: - Python...
This op seems to be used internally by `MelSpectrogram` layers and would be convenient to have supported. Potentially related: - https://github.com/tensorflow/tflite-micro/issues/1042
Hi, I want to build hello_world example for arm_cortex_m4 using "make -f tensorflow/lite/micro/tools/make/Makefile TARGET=cortex_m_corstone_300 TARGET_ARCH=cortex-m4 OPTIMIZED_KERNELS=cmsis_nn hello_world" command. But there is an error "section `.data' can't be allocated in segment...
I've searched through [schema_generated.h](https://github.com/tensorflow/tflite-micro/blob/main/tensorflow/lite/schema/schema_generated.h) and it seems like REVERSE_V2 is given Op Code 105 and included in the BuiltInOperator enumerations, however there's no reference implementation provided in [/micro/kernels](https://github.com/tensorflow/tflite-micro/tree/main/tensorflow/lite/micro/kernels). Without REVERSE_V2,...
This PR adds a few improvements to the `MicroProfiler` class. bug=#2725 First, I made it a template so that it's possible to adjust the maximum number of events the profiler...
PACK/UNPACK should support int16_t for int16x8 quantized models
The `MicroProfiler` class has these buffers: ```cpp const char* tags_[kMaxEvents]; uint32_t start_ticks_[kMaxEvents]; uint32_t end_ticks_[kMaxEvents]; struct TicksPerTag { const char* tag; uint32_t ticks; }; TicksPerTag total_ticks_per_tag[kMaxEvents]; ``` We need 20 bytes...
Currently a log is only printed for `kTfLiteError` status codes.
When trying to deploy the model on ESP32S3, I get a warning called `SUB failed to invoke INT32 not supported`. I have fixed it and I want to contribute by...
Metadata
Owner
Metadata
Infrastructure to enable deployment of ML models to low-power resource-constrained embedded targets (including microcontrollers and digital signal processors).