adds code


The video represents an audio input to EleutherAI GPT-J for text generation in Human-Robot interaction for making humanoid robots teaching assistants by using EleutherAI GPT-J-6B model and cloud-based NLP services. The aproach can be applied to EleutherAI GPT-J-NeoX and to all OpenAI GPT models that can use curl commands or Python APIs to interact with GPT. The Node-RED flow runs smoothly and could be applied by analogy for different NLP services provided by NLPcloud, such as Question answering in a piece of text, Summarizing text and Sentiment analysis. Temperature and Top-P are parameters of the model that can be set. When the Temperature and Top-P are set to zero the answers of the GPT-J is normal, while a high Temperature and high Top-P mean strange, more surprising choice of words as in the second response to the same question. More details how to develop the presented model for text generation in Node-RED can be found in: A. Lekova, P. Tsvetkova, T. Tanev, P. Mitrouchev and S. Kostova , "Making humanoid robots teaching assistants by using AI GPT-J model and cloud-based NLP services", accepted to Journal of Mechatronics and Artificial Intelligence in Engineering, JVE 2022
Previous Post Next Post