Fix refactoring params embeddings

src/training/train_triplet.py

@@ -96,7 +96,8 @@ def __init__(self, config_file):
 
             # Create optimizer with adaptive learning rate
             if self.config['freeze_convnet']:
-                params_names_embeddings = [f'embeddings.{j}.weight' for j in range(7)] + [f'embeddings.{j}.bias' for j in range(7)]
+                # params_names_embeddings = [f'embeddings.{j}.weight' for j in range(7)] + [f'embeddings.{j}.bias' for j in range(7)]
+                params_names_embeddings = ['embedding_layer.1.weight', 'embedding_layer.1.bias']
                 params_pt = [param for name, param in self.model.named_parameters() if name not in params_names_embeddings]
                 params_embeddings = [param for name, param in self.model.named_parameters() if name in params_names_embeddings]                 
                 # Overwrite optimizer
@@ -485,4 +486,4 @@ def get_nmr_embeddings(self):
     # Use this function to check if cdist gives the same result
     def euclidean_dist(self, emb_a, emb_b):
         dist = np.sqrt(np.dot(emb_a - emb_b, (emb_a - emb_b).T))
-        return dist
+        return dist