svmtrain function is returning a zero value when used in cross validation in libsvm

I am a student with a little basics in programming. But I had written a code in matlab for support vector regression using libsvm library and in it I am using cross validation for parameter optimization. But when I am running the code its returning a zero value for "svmtrain" function which is used in the cross validation part. I cant find any error in syntax but still the error is there. Please help me with this.

I am giving my code here. Please respond....

%Reading Input and Output Files

label = xlsread('label.xlsx');

instance = xlsread('instance.xlsx');

%Dividing Data for training and testing

train_label = label(1:76,:);

test_label = label(77:108,:);

train_instance = instance(1:76,:);

test_instance = instance(77:108,:);

%Cross Validation Search for finding best parameters

bestcv =0;

bestc=0;

bestg=0;

for log2c = -6:1:6,

for log2g = -6:1:6,

cmd = [' -v 4 -c ', num2str(2^log2c), ' -g ', num2str(2^log2g)];

        cv = svmtrain(train_label, train_instance, cmd)

        if (cv >= bestcv),
      bestcv = cv;
      bestc = 2^log2c;
      bestg = 2^log2g;
        end
        fprintf('%g %g %g (best c=%g, g=%g, rate=%g)\n', log2c, log2g, cv, bestc, bestg, bestcv);
  end
end

%Training the data set

options = ['-s 3 -t 2 -c ', num2str(bestc), ' -g ', num2str(bestg), '-p' 0.1, '-h' 0'];

model = svmtrain(train_label, train_instance, options);

%Testing the data set

[predicted_trainmodel, accuracy, decision_values] = svmpredict(train_label, train_instance, model);

[predicted_testmodel, accuracy, decision_values] = svmpredict(test_label, test_instance, model);