4 ms·
I have a experimental code that converts saved Keras deep learning models to a Racket friendly format, and I wrote a small Racket runtime for running pretrained
by mark_l_watson 8y ago
I have a experimental code that converts saved Keras deep learning models to a Racket friendly format, and I wrote a small Racket runtime for running pretrained models. My idea is treating models functionally and mixing in with old-School symbolic AI code. I am having some problems making the runtime efficient enough even using the native matrix libraries.