LINK TO THE geniter REPOSITORY – AI-generated imaginary letters, and more!
Doing the internet courses was completely manageable on my free time from work and family. However, when I wanted to do a quick prototype of an idea i had, i sadly realized it wasn’t at all as quick as i wanted it to be. I was lacking the experience and tools to do it.
So I took some time off work to immerse myself in machine learning, focusing on applying my knowledge and building a solid platform for prototyping my AI-ideas. Now, after 200+ hours of coding, i feel like I’ve come a long way.
The basic foundation for my AI-platform is:
- I work almost exclusively within Google Cloud, getting myself prepared to gear up when i need serious GPU-power
- I use Jupyter notebooks. Got used to them doing the Deep Learning specialization and i like the communicative capabilities.
- I use a combination of Tensorflow and Keras, taking advantage of the simplicity of Keras and the control of Tensorflow.
- I use git for version control, and Github for hosting. It’s the obvious choice, but i never took the time to implement it properly before.