fix comments

Author: apaszke

torch/optim/lbfgs.py

@@ -16,7 +16,7 @@ class LBFGS(Optimizer):
 
  .. note::
  This is a very memory intensive optimizer (it requires additional
- ``param_bytes * history_size`` bytes). If it doesn't fit in memory
+ ``param_bytes * (history_size + 1)`` bytes). If it doesn't fit in memory
  try reducing the history size, or use a different algorithm.
 
  Arguments:
@@ -73,7 +73,6 @@ def step(self, closure):
  closure (callable): A closure that reevaluates the model
  and returns the loss.
  """
- # TODO: backward after closure?
  assert len(self.param_groups) == 1
 
  group = self.param_groups[0]
@@ -156,7 +155,7 @@ def step(self, closure):
  al = state['al']
 
  for i in range(num_old):
- ro[i] = 1 / old_stps[i].dot(old_dirs[i])
+ ro[i] = 1. / old_stps[i].dot(old_dirs[i])
 
  # iteration in L-BFGS loop collapsed to use just one buffer
  q = flat_grad.neg()
@@ -189,7 +188,7 @@ def step(self, closure):
 
  # reset initial guess for step size
  if state['n_iter'] == 1:
- t = min(1, 1 / abs_grad_sum) * lr
+ t = min(1., 1. / abs_grad_sum) * lr
  else:
  t = lr