Reading training examples...done
Training set properties: 23 features, 90 rankings, 8174 examples
NOTE: Adjusted stopping criterion relative to maximum loss: eps=0.482200
Iter 1: .........*(NumConst=1, SV=1, CEps=482.2000, QPEps=0.0000)
Iter 2: .........*(NumConst=2, SV=2, CEps=445.2400, QPEps=0.0000)
Iter 3: .........*(NumConst=3, SV=2, CEps=29.7046, QPEps=0.0000)
Iter 4: .........*(NumConst=4, SV=2, CEps=18.4308, QPEps=2.8726)
Iter 5: .........*(NumConst=5, SV=2, CEps=2.9174, QPEps=0.9248)
Iter 6: .........*(NumConst=6, SV=2, CEps=0.5572, QPEps=0.0000)
Iter 7: .........*(NumConst=7, SV=3, CEps=0.7242, QPEps=0.2563)
Iter 8: .........(NumConst=7, SV=3, CEps=0.4345, QPEps=0.2563)
Final epsilon on KKT-Conditions: 0.43452
Upper bound on duality gap: 0.00004
Dual objective value: dval=0.03572
Primal objective value: pval=0.03577
Total number of constraints in final working set: 7 (of 7)
Number of iterations: 8
Number of calls to 'find_most_violated_constraint': 720
Number of SV: 3 
Norm of weight vector: |w|=0.05378
Value of slack variable (on working set): xi=343.01353
Value of slack variable (global): xi=343.19174
Norm of longest difference vector: ||Psi(x,y)-Psi(x,ybar)||=3237.06075
Runtime in cpu-seconds: 0.02
Compacting linear model...done
Writing learned model...done