model.py

import torch
import torch.nn as nn

class VotePredictor(nn.Module):
    def __init__(self, text_dim=384, country_count=193, country_emb_dim=32, hidden_dim=256):
        super(VotePredictor, self).__init__()
        self.country_embedding = nn.Embedding(country_count, country_emb_dim)
        self.model = nn.Sequential(
            nn.Linear(text_dim + country_emb_dim, hidden_dim),
            nn.ReLU(),
            nn.Dropout(0.3),
            nn.Linear(hidden_dim, 1)
        )

    def forward(self, text_vecs, country_ids):
        country_vecs = self.country_embedding(country_ids)
        x = torch.cat([text_vecs, country_vecs], dim=1)
        return self.model(x)