We have a robot war kind of event coming up next year. It’s between high-end bot developers and similar companies. As the lead of my team, I was wondering if there would be an option to not limit the hardware that can be used. What I had in mind to do is to use OpenCV and TensorFlow to train a model (on my laptop to take advantage of the NVIDIA GPU) or maybe even use a pre-trained model, not sure yet. Also, I want the bot to be smart and interactive, with the integration of sentiment analysis and NLP. For this we are seeking help from a chatbot development services (providing the link because I request anyone here to help understand if they are the right provider for what I’m looking for or not) provider. Combining both this software and hardware can we then deploy the model on any option we choose (Pi 4, Jetson Nano, etc)? And in case we decide to use an OAK camera or an Intel NCS could I later use OpenVINO to optimize and deploy the model that was trained with TensorFlow.
Hello @rvphilip!
I’m not familiar with the provider you mentioned but I believe that if what you want to check is the compatibility with different hardware the best option would be to ask them directly.
Since you mentioned TensorFlow I thought I should mention they have available multiple pre-trained models that you could check out:
https://tfhub.dev/s?module-type=text-language-model
Now, regarding the OpenVINO compatibility, I suggest checking this:
https://docs.openvino.ai/latest/openvino_docs_MO_DG_prepare_model_convert_model_Convert_Model_From_TensorFlow.html
https://docs.openvino.ai/latest/openvino_docs_MO_DG_prepare_model_convert_model_tf_specific_Convert_BERT_From_Tensorflow.html
I hope that helps!