davidsandberg / facenet

Face recognition using Tensorflow
MIT License
13.74k stars 4.8k forks source link

ValueError: The number of classes has to be greater than one; got 1 #1211

Open raj2022 opened 3 years ago

raj2022 commented 3 years ago

Hello, I am trying to do this code but I am obtaining this error, but do not know where to make changes. Please help me.

trainin gusing scikit learn

npy_data = np.load('back.npy')

Split numpy array into X (data) and Y (truth)

Y is just last column

Y = npy_data[:,-1]

X is all but last column

X = npy_data[:,:-1]

Random number list for splitting training and testing

r = np.random.rand(X.shape[0])

Get training and testing samples, splitting in half (using 0.5)

Xtrain=X[r>0.5] Xtest=X[r<=0.5] Ytrain=Y[r>0.5] Ytest=Y[r<=0.5]

Create and run the SVC with an rbf kernel

Some tuning has been done to get the gamma and C values.

method_sklearn = SVC(C = 2.0, kernel = 'rbf',tol=0.001,gamma=0.005) method_sklearn.fit(Xtrain,Ytrain)