psuedocode -> c_code

current best at this - also last model trained by me for this task. Following in the series go into english and french, then multimodal between images, audio, video, etc.