Unify gaussian likelihoods (#104)

* Unify gaussian likelihoods

* remove print

src/pythae/models/adversarial_ae/adversarial_ae_model.py

@@ -154,7 +154,7 @@ def loss_function(self, recon_x, x, z, z_prior):
 
         if self.model_config.reconstruction_loss == "mse":
 
-            recon_loss = F.mse_loss(
+            recon_loss = 0.5 * F.mse_loss(
                 recon_x.reshape(x.shape[0], -1),
                 x.reshape(x.shape[0], -1),
                 reduction="none",

src/pythae/models/beta_tc_vae/beta_tc_vae_model.py

@@ -92,7 +92,7 @@ def loss_function(self, recon_x, x, mu, log_var, z, dataset_size):
 
         if self.model_config.reconstruction_loss == "mse":
 
-            recon_loss = F.mse_loss(
+            recon_loss = 0.5 * F.mse_loss(
                 recon_x.reshape(x.shape[0], -1),
                 x.reshape(x.shape[0], -1),
                 reduction="none",

src/pythae/models/beta_vae/beta_vae_model.py

@@ -84,7 +84,7 @@ def loss_function(self, recon_x, x, mu, log_var, z):
 
         if self.model_config.reconstruction_loss == "mse":
 
-            recon_loss = F.mse_loss(
+            recon_loss = 0.5 * F.mse_loss(
                 recon_x.reshape(x.shape[0], -1),
                 x.reshape(x.shape[0], -1),
                 reduction="none",

src/pythae/models/ciwae/ciwae_model.py

@@ -95,7 +95,7 @@ def loss_function(self, recon_x, x, mu, log_var, z):
 
         if self.model_config.reconstruction_loss == "mse":
 
-            recon_loss = F.mse_loss(
+            recon_loss = 0.5 * F.mse_loss(
                 recon_x,
                 x.reshape(recon_x.shape[0], -1)
                 .unsqueeze(1)

src/pythae/models/disentangled_beta_vae/disentangled_beta_vae_model.py

@@ -92,7 +92,7 @@ def loss_function(self, recon_x, x, mu, log_var, z, epoch):
 
         if self.model_config.reconstruction_loss == "mse":
 
-            recon_loss = F.mse_loss(
+            recon_loss = 0.5 * F.mse_loss(
                 recon_x.reshape(x.shape[0], -1),
                 x.reshape(x.shape[0], -1),
                 reduction="none",

src/pythae/models/factor_vae/factor_vae_model.py

@@ -137,7 +137,7 @@ def loss_function(self, recon_x, x, mu, log_var, z, z_bis_permuted):
 
         if self.model_config.reconstruction_loss == "mse":
 
-            recon_loss = F.mse_loss(
+            recon_loss = 0.5 * F.mse_loss(
                 recon_x.reshape(x.shape[0], -1),
                 x.reshape(x.shape[0], -1),
                 reduction="none",

src/pythae/models/info_vae/info_vae_model.py

@@ -92,7 +92,7 @@ def loss_function(self, recon_x, x, z, z_prior, mu, log_var):
 
         if self.model_config.reconstruction_loss == "mse":
 
-            recon_loss = F.mse_loss(
+            recon_loss = 0.5 * F.mse_loss(
                 recon_x.reshape(x.shape[0], -1),
                 x.reshape(x.shape[0], -1),
                 reduction="none",

src/pythae/models/iwae/iwae_model.py

@@ -94,7 +94,7 @@ def loss_function(self, recon_x, x, mu, log_var, z):
 
         if self.model_config.reconstruction_loss == "mse":
 
-            recon_loss = F.mse_loss(
+            recon_loss = 0.5 * F.mse_loss(
                 recon_x,
                 x.reshape(recon_x.shape[0], -1)
                 .unsqueeze(1)

src/pythae/models/miwae/miwae_model.py

@@ -103,7 +103,7 @@ def loss_function(self, recon_x, x, mu, log_var, z):
 
         if self.model_config.reconstruction_loss == "mse":
 
-            recon_loss = F.mse_loss(
+            recon_loss = 0.5 * F.mse_loss(
                 recon_x,
                 x.reshape(recon_x.shape[0], -1)
                 .unsqueeze(1)

src/pythae/models/piwae/piwae_model.py

@@ -111,7 +111,7 @@ def loss_function(self, recon_x, x, mu, log_var, z):
 
         if self.model_config.reconstruction_loss == "mse":
 
-            recon_loss = F.mse_loss(
+            recon_loss = 0.5 * F.mse_loss(
                 recon_x,
                 x.reshape(recon_x.shape[0], -1)
                 .unsqueeze(1)

src/pythae/models/pvae/pvae_model.py

@@ -128,7 +128,7 @@ def loss_function(self, recon_x, x, z, qz_x):
 
         if self.model_config.reconstruction_loss == "mse":
 
-            recon_loss = F.mse_loss(
+            recon_loss = 0.5 * F.mse_loss(
                 recon_x.reshape(x.shape[0], -1),
                 x.reshape(x.shape[0], -1),
                 reduction="none",

src/pythae/models/svae/svae_model.py

@@ -101,7 +101,7 @@ def loss_function(self, recon_x, x, loc, concentration, z):
 
         if self.model_config.reconstruction_loss == "mse":
 
-            recon_loss = F.mse_loss(
+            recon_loss = 0.5 * F.mse_loss(
                 recon_x.reshape(x.shape[0], -1),
                 x.reshape(x.shape[0], -1),
                 reduction="none",

src/pythae/models/vae/vae_model.py

@@ -100,8 +100,7 @@ def forward(self, inputs: BaseDataset, **kwargs):
     def loss_function(self, recon_x, x, mu, log_var, z):
 
         if self.model_config.reconstruction_loss == "mse":
-
-            recon_loss = F.mse_loss(
+            recon_loss = 0.5 * F.mse_loss(
                 recon_x.reshape(x.shape[0], -1),
                 x.reshape(x.shape[0], -1),
                 reduction="none",

src/pythae/models/vae_gan/vae_gan_model.py

@@ -192,7 +192,7 @@ def loss_function(self, recon_x, x, z, z_prior, mu, log_var):
         )[f"embedding_layer_{self.reconstruction_layer}"]
 
         # MSE in feature space
-        recon_loss = F.mse_loss(
+        recon_loss = 0.5 * F.mse_loss(
             true_discr_layer.reshape(N, -1),
             recon_discr_layer.reshape(N, -1),
             reduction="none",

src/pythae/models/vae_iaf/vae_iaf_model.py

@@ -106,7 +106,7 @@ def loss_function(self, recon_x, x, mu, log_var, z0, zk, log_abs_det_jac):
 
         if self.model_config.reconstruction_loss == "mse":
 
-            recon_loss = F.mse_loss(
+            recon_loss = 0.5 * F.mse_loss(
                 recon_x.reshape(x.shape[0], -1),
                 x.reshape(x.shape[0], -1),
                 reduction="none",

src/pythae/models/vae_lin_nf/vae_lin_nf_model.py

@@ -119,7 +119,7 @@ def loss_function(self, recon_x, x, mu, log_var, z0, zk, log_abs_det_jac):
 
         if self.model_config.reconstruction_loss == "mse":
 
-            recon_loss = F.mse_loss(
+            recon_loss = 0.5 * F.mse_loss(
                 recon_x.reshape(x.shape[0], -1),
                 x.reshape(x.shape[0], -1),
                 reduction="none",

src/pythae/models/vamp/vamp_model.py

@@ -104,7 +104,7 @@ def loss_function(self, recon_x, x, mu, log_var, z, epoch):
 
         if self.model_config.reconstruction_loss == "mse":
 
-            recon_loss = F.mse_loss(
+            recon_loss = 0.5 * F.mse_loss(
                 recon_x.reshape(x.shape[0], -1),
                 x.reshape(x.shape[0], -1),
                 reduction="none",