Manas Mohanty

Hi @Raunak-Singh-Inventor ! You can train the model through custom image dataset and get the respective C++ files for including in the Arduino librarary using [Edge impulse ](https://www.edgeimpulse.com/). Could you...

Hi @Raunak-Singh-Inventor ! I was able to use edge-impulse with my personal account too. You can also refer this [document](https://www.tensorflow.org/lite/microcontrollers/get_started_low_level#run_inference) for writing the loading and inference part after getting .h...

Ok @Raunak-Singh-Inventor ! Thanks for clarifying your requirements. Importing an Tensorflow model does not seem to be part of feature yet in Edge impulse ( But the features like Data...

Hi @rolfvandam ! Thanks for reporting this bug. It is replicating even when a valid path is given . @rthadur ! Could you look at this issue . Attached gist...

Hi @rolfvandam ! Please Check this [comment](https://github.com/tensorflow/tensorflow/issues/57386#issuecomment-1223836276) and resolved [gist](https://colab.sandbox.google.com/gist/mohantym/20aef7edda23af1fdb655350c1d96de6/model-maker-object-detection-tutorial.ipynb#scrollTo=qhl8lqVamEty) as a workaround . Raised PR [#57394 ](https://github.com/tensorflow/tensorflow/pull/57394)too upon this. Thank you!

Hi @misterBart ! Thanks for bring it up . @sachinprasadhs ! Could you look at these issues. Thank you!

@code4lala ! Thanks for pointing it out. @sachinprasadhs ! Could you look at this doc-bug. Thank you!

Hi @wikstorm ! 1. Could you set a path as below. `export XLA_FLAGS=--xla_gpu_cuda_data_dir= ` 2. If issue still persists, Please delete existing installation and switch to Cuda 11.2 and CudNN...

Ok @wikstorm ! Thanks for the update. I also suggest to check step 2 in above [comment](https://github.com/tensorflow/tensorflow/issues/57055#issuecomment-1210272066) for windows machine. Please let us know whether it is resolved or not...

@slavazhulanov ! Check this [thread](https://github.com/tensorflow/tensorflow/issues/56927#issuecomment-1236291285) once.