System-Optimization-For-AI
Most of us do not think about why we tire the system when dealing with data science, machine learning, deep learning. However, if we can use our system at full performance, it will positively affect the work we do. It is also possible to do the same job with less RAM, Core, GPU! All you need to do is to use your system in a more optimized way. So how will this happen?
Example_1: You have a Deep Learning model and big data. As a result, your RAM and VRAM needs will be high. You can see your system's values in each epoch and update your project accordingly. You can use "Swap_Memory" for RAM and functions that allow you to free your VRAM in every for loop for VRAM.
Example_2: Let's say you have a large data set. But you have to work separately. If you have 32 threads, it means you can assign 16 threads to both data groups. By doing this, you will get more stable, efficient and faster results than starting 2 codes at the same time.
Example_3: Observing the system while it is on the job and detecting defective parts.
Examples can be multiplied further. It will be beneficial for you and your system to use the functions here by designing them according to your purpose and imagination.
Here we have explained the libraries needed to optimize your system.