using no cache
Reading from train-sets/0002.dat
num sources = 1
Num weight bits = 18
learning rate = 10
initial_t = 1
power_t = 0.5
predictions = 0002c.predict
only testing
average    since       example  example    current  current  current
loss       last        counter   weight      label  predict features
0.002276   0.002276          3      3.0     0.5498   0.5361      184
0.031357   0.060437          6      6.0     0.2681   0.6668      184
0.027939   0.023839         11     11.0     0.4315   0.6007      184
0.042069   0.056199         22     22.0     0.5519   0.4997      184
0.025585   0.009101         44     44.0     0.5514   0.5331      184
0.024013   0.022405         87     87.0     0.5140   0.5308      184
0.018521   0.013029        174    174.0     0.5596   0.4972      184
0.016111   0.013701        348    348.0     0.5475   0.4388      184
0.015105   0.014098        696    696.0     0.3421   0.7898      184
0.014822   0.014539       1392   1392.0     0.4996   0.5059      184
0.013016   0.011211       2784   2784.0     0.5090   0.3869      184
0.012228   0.011440       5568   5568.0     0.6413   0.7550      184
0.011123   0.010018      11135  11135.0     0.3869   0.4848      184
0.011121   0.011119      22269  22269.0     0.5063   0.4510      184
0.015806   0.020491      44537  44537.0     0.4905   0.5220      184

finished run
number of examples = 74746
weighted example sum = 6.952e+04
weighted label sum = 3.511e+04
average loss = 0.01421
best constant = 0.5051
best constant's loss = 0.25
total feature number = 13753264