[Feat] : Add AdaGrad, RMSProp, and Adam Optimizers

[Advaitgaur004]

Just added a few more optimisers to have some more powerful options for training! This PR brings in AdaGrad, RMSProp, and the classic Adam.

• Added the new optimizers (AdaGrad, RMSProp, Adam) to the main optimizer file. I put them all in src/optimizer/sgd.c for now, but we can totally break them into separate files later if we want.
• Updated cten.h with the new function declarations so they can be used across the project.
• Kept SGD as is, so nothing breaks there.

[Advaitgaur004]

to test the implemenation use this ==> https://gist.github.com/Advaitgaur004/e0ad0e99f1e4514bd6091f3cce10b23a

[Advaitgaur004]

As expected, AdaGrad diminishes the learning rate, and after 300 iterations, it is about to reach the destination, but not entirely.

[Advaitgaur004]

Choosing an optimizer can be tricky....many people default to Adam, but sometimes SGDM performs better.
Refer to this --> Here

[Advaitgaur004]

Extracted symbols from: https://gist.github.com/mrquincle/1102e58562411b7f633c08d5dc9a2e1f