Added Adam optimizer implementation

[Spnetic-5]

Adam Implementation:

• Momentum Update:
  m = beta1 * m + (1 - beta1) * gradient

• Adaptive Learning Rates Update:
  v = beta2 * v + (1 - beta2) * gradient**2

• Bias-Corrected First and Second Moment Estimates:
  m_hat = m / (1 - beta1**t)
v_hat = v / (1 - beta2**t)

• Parameter Update:
  param = param - learning_rate * m_hat / (sqrt(v_hat) + epsilon)

[milancurcic]

Thank you! Good so far.

[milancurcic]

Thank you @Spnetic-5, this is almost good to go. The only thing missing is the convergence test for Adam in test/test_optimizers.f90.

[milancurcic]

CI seems to have trouble downloading test files from GitHub. Not real failures; tests pass on my end.

[milancurcic]

@Spnetic-5, just checking, your latest commit message mentions the convergence test but the commit doesn't actually include the test.

[Spnetic-5]

@Spnetic-5, just checking, your latest commit message mentions the convergence test but the commit doesn't actually include the test.

Sorry, forgot to push test file😅️

[milancurcic]

Thank you, @Spnetic-5, will merge. Please release v0.14.0 when ready.