your learning rate is too high, setting it to 1
Num weight bits = 13
learning rate = 1
initial_t = 1
power_t = 0.5
using no cache
Reading from train-sets/wiki1K.dat
num sources = 1
average    since         example     example  current  current  current
loss       last          counter      weight    label  predict features
10.312708  10.312708           3         3.0  unknown   0.0000       37
10.402514  10.492320           6         6.0  unknown   0.0000       13
10.356154  10.300523          11        11.0  unknown   0.0000       31
11.649708  12.943261          22        22.0  unknown   0.0000        1
11.039215  10.428723          44        44.0  unknown   0.0000      165
11.042017  11.044884          87        87.0  unknown   0.0000       28
10.370092  9.698167          174       174.0  unknown   0.0000       16
9.628472   8.886853          348       348.0  unknown   0.0000        1
8.989231   8.349991          696       696.0  unknown   0.0000      142

finished run
number of examples = 1000
weighted example sum = 1000
weighted label sum = 0
average loss = 8.742
best constant = 1
total feature number = 86919