Num weight bits = 18
learning rate = 10
initial_t = 1
power_t = 0.5
decay_learning_rate = 1
creating cache_file = train-sets/multiclass.cache
Reading from train-sets/multiclass
num sources = 1
average    since         example     example  current  current  current
loss       last          counter      weight    label  predict features
0.666667   0.666667          3      3.0          3        2        2
0.833333   1.000000          6      6.0          6        5        2
0.909091   1.000000         11     11.0          1       10        2
0.500000   0.090909         22     22.0          2        2        2
0.250000   0.000000         44     44.0          4        4        2
0.126437   0.000000         87     87.0          7        7        2

finished run
number of examples = 100
weighted example sum = 100
weighted label sum = 0
average loss = 0.11
best constant = -0.0101
total feature number = 200