We are now ready for training our text classifier model. Let's start with something simple: we are going to consider this model to be a black box for now.
Model architecture is better explained by other sources, including several YouTube videos such as those by CS224n at Stanford (http://web.stanford.edu/class/cs224n/). I suggest that you explore and connect it with the know-how that you already have:
class SimpleLSTMBaseline(nn.Module): def __init__(self, hidden_dim, emb_dim=300, spatial_dropout=0.05, recurrent_dropout=0.1, num_linear=2): super().__init__() # don't forget to call this! self.embedding = nn.Embedding(len(TEXT.vocab), emb_dim) self.encoder = nn.LSTM(emb_dim, hidden_dim, num_layers=num_linear, ...
