Num weight bits = 18
learning rate = 0.5
initial_t = 0
power_t = 0.5
final_regressor = models/bs.reg.model
predictions = bs.reg.predict
using no cache
Reading datafile = train-sets/0002.dat
num sources = 1
average    since         example     example  current  current  current
loss       last          counter      weight    label  predict features
0.271591   0.271591            1         1.0   0.5211   0.0000       15
0.218951   0.166311            2         2.0   0.5353   0.1274       15
0.133586   0.048220            4         4.0   0.5854   0.2751       15
0.110732   0.087879            8         8.0   0.5575   0.3162       15
0.096665   0.082598           16        16.0   0.5878   0.3767       15
0.056579   0.016494           32        32.0   0.6038   0.6028       15
0.042532   0.028485           64        64.0   0.5683   0.3916       15
0.029754   0.016976          128       128.0   0.5351   0.4394       15
0.017927   0.006101          256       256.0   0.5385   0.5544       15
0.010301   0.002674          512       512.0   0.5053   0.5092       15
0.005649   0.000997         1024      1024.0   0.5750   0.5904       15
0.003118   0.000588         2048      2048.0   0.5204   0.4896       15
0.001697   0.000276         4096      4096.0   0.5042   0.4924       15
0.001104   0.000511         8192      8192.0   0.4967   0.5365       15
0.000779   0.000453        16384     16384.0   0.5011   0.4998       15
0.000678   0.000578        32768     32768.0   0.3915   0.4169       15
0.000807   0.000937        65536     65536.0   0.5043   0.4820       15

finished run
number of examples per pass = 74746
passes used = 1
weighted example sum = 69521
weighted label sum = 0
average loss = 0.000818056
best constant = 0
total feature number = 1119986