ValueError: The `kernel_size` argument must be a tuple of 1 integers when use params for kernel size

First off, make sure to check your support options.

The preferred way to resolve usage related matters is through the docs which are maintained up-to-date with the latest version of Talos.

If you do end up asking for support in a new issue, make sure to follow the below steps carefully.

1) Confirm the below

[X] I have looked for an answer in the Docs
[X] My Python version is 3.5 or higher
[X] I have searched through the issues Issues for a duplicate
[X] I've tested that my Keras model works as a stand-alone

2) Include the output of:

talos.__version__

1.0.0

3) Explain clearly what you are trying to achieve

Dear Developers, I want to use Talos with my pre-defined validation set. Let say i have three dataframe which i call Train, Test and Validation. I want to use my Validation dataset as validation for my model beside let talos make it for me.

4) Explain what you have already tried

I have try to use Talos in a separate function and run it. It gives me error: ValueError: The `kernel_size` argument must be a tuple of 1 integers. Received: [3, 5, 7]. I use Conv1D which is differ and not found any case with this in Talos Docs.

5) Provide a code-complete reference

This is how my params are made:

dropout_rate = [0.2, 0.5, 0.7]
feature_len = [64,128,256]
wconv = [3,5,7]
epc = [100,200,500]
para = {'dropout_rate':dropout_rate,'feature_len':feature_len,'wconv':wconv,'epc':epc}

This is how i make the model:

def GetModel(trainX,trainY,validX,validY,params):
   def GetModel(trainX,trainY,x_val,y_val,params):
    model = tf.keras.Sequential([
       Conv1D(filters=params['feature_len'],kernel_size=params['wconv'],activation='relu'),
        Dropout(params['dropout_rate']),
        MaxPooling1D(pool_size=2,strides=1),
        Flatten(),
        Dense(2,activation='softmax')
    ])
    model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy','mse'])
    out=model.fit(x=trainX,
                  y=trainY,
                  epochs=params['epc'],
                  validation_data=[x_val, y_val])
    return model,out

This is how i put talos in a different function:

def training(trainX,validX,trainY,validY,params):
    model,out = GetModel(trainX,validX,trainY,validY,params)
    scan_object = talos.Scan(x=trainX,
                             y=trainY, 
                             model=GetModel, 
                             x_val=validX,
                             y_val=validY, 
                             params=para, 
                             experiment_name='ECG')
    return model,scan_object,out

This is how i run my model with a 100 times test:

def experimental_test(n_exp,sddb,nsr,n_train,n_valid,n_test,start_normal,start_sddb,options):
       ....
        train = [A_train,B_train]
        valid = [A_val,B_val]
        tests = [A_test,B_test]
        train_dt = pd.concat(train)
        valid_dt = pd.concat(valid)
        test_dt = pd.concat(tests)
        trainX,validX,trainY,validY= dataset_maker(train_dt,valid_dt)
        n_timesteps, n_features, n_outputs =trainX.shape[1], trainX.shape[2], trainY.shape[0]
        model,scan_object,out=training(trainX,validX,trainY,validY,para)
        pyplot.plot(out.history['loss'], label='train')
        pyplot.plot(out.history['val_loss'], label='validation')
        pyplot.legend()
        pyplot.show()
        pyplot.plot(out.history['accuracy'], label='train')
        pyplot.plot(out.history['val_accuracy'], label='validation')
        pyplot.legend()
        pyplot.show()
        test_seq(start_sddb,start_normal,test_dt,model)

[ ] My support question includes an input model
[X] My support question includes a parameter dictionary
[X] My support question includes a Scan() command
[ ] My support question includes a link to a sample of the data

NOTE: If the data is sensitive and can't be shared, create dummy data that mimics it.

A self-contained Jupyter Notebook, Google Colab, or similar is highly preferred and will speed up helping you with your issue.

autonomio / talos