Using a third variable list while designing custom loss functions in Keras

I am trying to create a loss function for a sports betting model. I am having trouble bringing the odds (a third variable) into the loss function. When I do bring it in, it works for the first batch but seems to revert back to the start of my list/tensor when a new batch begins. Is it possible to iterate though the odds list as rate as the yTrue/yPred values.

odds = data_info['FRD_Open'] # Specific value for each model prediction

def custom_Loss(odds):

     def customLoss(yTrue, yPred):

         o = tf.convert_to_tensor(odds, dtype=tf.float32)
         loss = K.mean(K.sum(K.relu((yTrue * (o - 1) - (1 - yTrue)) * yPred), axis=1))

         return -loss

     return customLoss

Note: data_info is a pandas dataframe of the data used to train and test the model.