Fix L-BFGS options

[vl-dud]

This PR fixes dde.optimizers.set_LBFGS_options so that it correctly updates iter_per_step and fun_per_step used by L-BFGS in PyTorch/Paddle backends.

Before PR:

dde.optimizers.set_LBFGS_options(maxiter=123, maxfun=456)
print(dde.optimizers.LBFGS_options)
# {'maxcor': 100, 'ftol': 0, 'gtol': 1e-08, 'maxiter': 123, 'maxfun': 456, 'maxls': 50, 'iter_per_step': 1000, 'fun_per_step': 1250}

After PR:

dde.optimizers.set_LBFGS_options(maxiter=123, maxfun=456)
print(dde.optimizers.LBFGS_options)
# {'maxcor': 100, 'ftol': 0, 'gtol': 1e-08, 'maxiter': 123, 'maxfun': 456, 'maxls': 50, 'iter_per_step': 123, 'fun_per_step': 456}

[echen5503]

Please verify my understanding: set_hvd_opt_options sets the LFBGS options along with doing the backend-specific operations, but in reality set_LFBGS_options should handle this itself, causing the bug that when you call set_LGBGS_options by itself, it wont update iter_per_step and fun_per_step.

[vl-dud]

set_LBFGS_options is expected to update all L-BFGS-related parameters, including iter_per_step and fun_per_step used in the Paddle and PyTorch backends. However, the current implementation does not update them.

[echen5503]

Ok. This looks good.