search

youxch / Inverse-design-of-patch-antennas

This repository hosts a simple demonstration of a deep learning approach for the inverse design of patch antennas. The goal is to explore energy-efficient designs and to significantly reduce simulation cost compared to conventional methods.
MIT License
48 stars 7 forks source link

Parameter issues #8

Open L-Zhe-o opened 1 month ago

L-Zhe-o commented 1 month ago

Thanks for sharing, I have a question about the parameters, what does the Initial parameter refer to in this image 321282217-9a992852-f7e6-4ca1-b9fc-a0a993c7fe81

youxch commented 1 month ago

Thank you for your question. The initial parameters refer to the structural parameters made by Professor K. L. Wong (Figure below). image

L-Zhe-o commented 1 month ago

Thanks for the reply, I see what the Initial parameter means. I also have another question about the code, when running the predict code is always repeating the train code running process, is this problem caused by my own mistakes? Or has it always existed? Looking forward to hearing from you.

youxch commented 1 month ago

Thank you for your question. The predict.py script should not invoke train.py during its process. I suspect you might have mistaken the output of the prediction process for the training process. The role of predict.py is to predict over 2 million structures at a rate of about one prediction per millisecond. If you wish to accelerate this prediction process in parallel, you can run predict-gpu.py, but you will need to set up the CUDA environment accordingly. I hope my answer is helpful to you. If you still have any issues, please feel free to post the runtime results, and we can discuss further. You can also refer to this link (maybe the same question) https://github.com/youxch/Inverse-design-of-patch-antennas/issues/2#issuecomment-2046543092