doc: add note about how to use tflm on nuclei riscv processor using n…

…sdk 0.3.8

Signed-off-by: Huaqi Fang <[email protected]>

README.md

@@ -19,6 +19,8 @@ TensorFlow Lite for Microcontrollers is a port of TensorFlow Lite designed to
 run machine learning models on DSPs, microcontrollers and other devices with
 limited memory.
 
+> **NOTE**: Looking for how to run on Nuclei RISC-V Processor, please check [this guidance](tensorflow/lite/micro/nuclei_demosoc/README.md)
+
 Additional Links:
  * [Tensorflow github repository](https://github.com/tensorflow/tensorflow/)
  * [TFLM at tensorflow.org](https://www.tensorflow.org/lite/microcontrollers)

tensorflow/lite/micro/nuclei_demosoc/README.md

@@ -4,7 +4,11 @@ This folder contains TFLite kernel operations optimized for Nuclei RISC-V Proces
 
 It is designed to be portable even to 'bare metal', so it follows the same design goals as the micro experimental port.
 
-## How to quick explore it
+# How to quick explore it
+
+This TFLM support is done based on Nuclei SDK 0.3.8 release, see https://github.com/Nuclei-Software/nuclei-sdk/releases/tag/0.3.8
+
+It mainly works on Nuclei Demo SoC, an fpga prototype SoC to test different kinds of Nuclei RISC-V Processor.
 
 ## Install third party requirements