Hemanth-Jonnala

can you please check the permissions of` /dev/apex_0` and check if [this](https://github.com/tensorflow/tensorflow/issues/35054#issuecomment-716477440) works for you.

can you please try the demo with the below lines [here](https://github.com/google-coral/pycoral/blob/master/examples/classify_image.py#L41) and share the logs.. ``` from pycoral.pybind._pywrap_coral import SetVerbosity as set_verbosity set_verbosity(10) ```

I am not sure how to fix the device open error.. Seems to be issue with MSI-X support (lspci -vvv|grep -i MSI-X).. Might be your host machine does not support...

@tedzhouhk can you please share the following details: ``` Python 3.9.7 (default, Sep 3 2021, 06:18:44) [GCC 10.2.1 20210110] on linux Type "help", "copyright", "credits" or "license" for more information....

@tedzhouhk please add these two lines to the [demo](https://github.com/google-coral/pycoral/blob/master/examples/classify_image.py#L41) and share the output in txt file. ``` from pycoral.pybind._pywrap_coral import SetVerbosity as set_verbosity set_verbosity(10) ```

can you try the Workaround to disable Apex and Gasket section form this page: https://coral.ai/docs/m2/get-started/#troubleshooting-on-linux

@tedzhouhk Yes, please try the Workaround to disable Apex and Gasket section. If still not working, if possible please try with Linux container or Ubuntu 18.04.

Hello @gworkman Yes, LSTM operation seems to have size limitation. Please try the below the options here: 1. Reduce the model input size 2. Map the LSTM op to CPU...

Closing this issue as we haven't find any concrete limits for input shape/LSTM number of units to document the limitations. Thanks!

@jiaqiyin1995 the issue is with Batch Mat Mul operation. It is a bug with the compiler as it is mapping the batch mat mul operation to edgeTPU even though it...