def fit(self, trX, trY, n_iter=1):
out_shape = self.model[-1].out_shape
n = 0.
t = time()
costs = []
for e in range(n_iter):
epoch_costs = []
for xmb, ymb in self.iterator.iterXY(trX, trY):
c = self._train(xmb, ymb)
epoch_costs.append(c)
n += len(ymb)
if self.verbose >= 2:
n_per_sec = n / (time() - t)
n_left = len(trY) * n_iter - n
time_left = n_left / n_per_sec
sys.stdout.write(
"\rIter %d Seen %d samples Avg cost %0.4f Examples per second %d Time left %d seconds"
% (e, n, np.mean(
epoch_costs[-250:]), n_per_sec, time_left))
sys.stdout.flush()
costs.extend(epoch_costs)
n_per_sec = n / (time() - t)
n_left = len(trY) * n_iter - n
time_left = n_left / n_per_sec
status = "Iter %d Seen %d samples Avg cost %0.4f Examples per second %d Time elapsed %d seconds" % (
e, n, np.mean(epoch_costs[-250:]), n_per_sec, time() - t)
if self.verbose >= 2:
sys.stdout.write("\r" + status)
sys.stdout.flush()
sys.stdout.write("\n")
elif self.verbose == 1:
print(status)
return costs
def fit(self, trX, trY, n_iter=1):
out_shape = self.model[-1].out_shape
n = 0.
t = time()
costs = []
for e in range(n_iter):
epoch_costs = []
for xmb, ymb in self.iterator.iterXY(trX, trY):
c = self._train(xmb, ymb)
epoch_costs.append(c)
n += len(ymb)
if self.verbose >= 2:
n_per_sec = n / (time() - t)
n_left = len(trY)*n_iter - n
time_left = n_left/n_per_sec
sys.stdout.write("\rIter %d Seen %d samples Avg cost %0.4f Examples per second %d Time left %d seconds" % (e, n, np.mean(epoch_costs[-250:]), n_per_sec, time_left))
sys.stdout.flush()
costs.extend(epoch_costs)
n_per_sec = n / (time() - t)
n_left = len(trY)*n_iter - n
time_left = n_left/n_per_sec
status = "Iter %d Seen %d samples Avg cost %0.4f Examples per second %d Time elapsed %d seconds" % (e, n, np.mean(epoch_costs[-250:]), n_per_sec, time() - t)
if self.verbose >= 2:
sys.stdout.write("\r"+status)
sys.stdout.flush()
sys.stdout.write("\n")
elif self.verbose == 1:
print(status)
return costs