Do I have to input to a model for the speaker recognition the amount of speakers? If I do will it cause problems when testing?

So I have this model with an encoder-discriminator system, not exactly a GAN though. Here are the fragments with the speaker number:

class Model(nn.Module):
    def __init__(self, num_speakers):
        super(Model, self).__init__()

        self.discriminator = Discriminator()
        self.encoder = Encoder()
        self.classifier = Classifier(num_speakers)

    def forward(self, input, speakers):
        print("input size", input.shape)
        input = input.reshape((BATCHES * 3, 1, 3200))
        embeddings = self.encoder(input.to("cuda:0"))  # TODO
        speakers_probs = self.classifier(embeddings.to("cuda:0"))
        embeddings = embeddings.reshape((BATCHES, 3, constants.embedding_size))
....


class Classifier(nn.Module):
    def __init__(self, num_speakers):
        super(Classifier, self).__init__()
        self.fc1 = nn.Linear(constants.embedding_size, constants.embedding_size) #inner layer
        self.proj = nn.Linear(constants.embedding_size, num_speakers)

    def encode(self, x):
        return Fun.relu(self.fc1(x))

    def forward(self, x):
        x = self.encode(x) #hidden representation
        x = Fun.softmax(self.proj(x), dim=1)
        return x
...


class Discriminator(nn.Module):
    def __init__(self):
        super(Discriminator, self).__init__()
        self.fc1 = nn.Linear(2 * constants.embedding_size, 2 * constants.embedding_size)
        self.proj = nn.Linear(2 * constants.embedding_size, 1)

    def forward(self, x):
        x = Fun.relu(self.fc1(x))
        x = self.proj(x)
        x = torch.sigmoid(x)
        return x

It feels awkward to pass it in the main as model = Model(866) (the amount of speakers), and I have doubts if it won't be a problem when using the model for testing if it will recognize a person when the amount of speakers will be 1.

source https://stackoverflow.com/questions/73338868/do-i-have-to-input-to-a-model-for-the-speaker-recognition-the-amount-of-speakers