model.py

import torch
import torch.nn as nn
import torch.nn.functional as F
from transformers import PreTrainedModel, PretrainedConfig


class BaseVAE(nn.Module):
    def __init__(self, latent_dim=16):
        super(BaseVAE, self).__init__()
        self.latent_dim = latent_dim

        self.encoder = nn.Sequential(
            nn.Conv2d(3, 32, 4, 2, 1),    # 32x32 -> 16x16
            nn.BatchNorm2d(32),
            nn.ReLU(),
            nn.Conv2d(32, 64, 4, 2, 1),   # 16x16 -> 8x8
            nn.BatchNorm2d(64),
            nn.ReLU(),
            nn.Conv2d(64, 128, 4, 2, 1),  # 8x8 -> 4x4
            nn.BatchNorm2d(128),
            nn.ReLU(),
            nn.Flatten()
        )
        self.fc_mu = nn.Linear(128 * 4 * 4, latent_dim)
        self.fc_logvar = nn.Linear(128 * 4 * 4, latent_dim)

        self.decoder_input = nn.Linear(latent_dim, 128 * 4 * 4)
        self.decoder = nn.Sequential(
            nn.ConvTranspose2d(128, 64, 4, 2, 1),  # 4x4 -> 8x8
            nn.BatchNorm2d(64),
            nn.ReLU(),
            nn.ConvTranspose2d(64, 32, 4, 2, 1),   # 8x8 -> 16x16
            nn.BatchNorm2d(32),
            nn.ReLU(),
            nn.ConvTranspose2d(32, 3, 4, 2, 1),    # 16x16 -> 32x32
            nn.Sigmoid()
        )

    def encode(self, x):
        x = self.encoder(x)
        mu = self.fc_mu(x)
        logvar = self.fc_logvar(x)
        return mu, logvar

    def reparameterize(self, mu, logvar):
        std = torch.exp(0.5 * logvar)
        eps = torch.randn_like(std)
        return mu + eps * std

    def decode(self, z):
        x = self.decoder_input(z)
        x = x.view(-1, 128, 4, 4)
        return self.decoder(x)

    def forward(self, x):
        mu, logvar = self.encode(x)
        z = self.reparameterize(mu, logvar)
        recon = self.decode(z)
        return recon, mu, logvar

class VAEConfig(PretrainedConfig):
    model_type = "vae"

    def __init__(self, latent_dim=16, **kwargs):
        super().__init__(**kwargs)
        self.latent_dim = latent_dim

class VAEModel(PreTrainedModel):
    config_class = VAEConfig

    def __init__(self, config):
        super().__init__(config)
        self.vae = BaseVAE(latent_dim=config.latent_dim)
        self.post_init()

    def forward(self, x):
        return self.vae(x)

    def encode(self, x):
        return self.vae.encode(x)

    def decode(self, z):
        return self.vae.decode(z)

device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
model = VAEModel.from_pretrained("BioMike/emoji-vae-init").to(device)
model.eval()