Merge pull request #1414 from TangJiakai/master

FwFMs Bug fix

recbole/model/context_aware_recommender/ffm.py

@@ -165,7 +165,7 @@ def __init__(self, feature_names, feature_dims, feature2id, feature2field, num_f
         # init float field-aware embeddings if there is float type of features.
         if len(self.float_feature_names) > 0:
             self.num_float_features = len(self.float_feature_names)
-            self.float_embeddings = nn.Embedding(np.sum(self.token_feature_dims, dtype=np.int32), self.embed_dim)
+            # self.float_embeddings = nn.Embedding(np.sum(self.token_feature_dims, dtype=np.int32), self.embed_dim)
             self.float_embeddings = torch.nn.ModuleList([
                 nn.Embedding(self.num_float_features, self.embed_dim) for _ in range(self.num_fields)
             ])

recbole/model/context_aware_recommender/fwfm.py

@@ -51,6 +51,8 @@ def __init__(self, config, dataset):
         self.num_fields = len(set(self.feature2field.values()))  # the number of fields
         self.num_pair = self.num_fields * self.num_fields
 
+        self.weight = torch.randn(self.num_fields,self.num_fields,1,requires_grad=True,device=self.device)
+
         self.loss = nn.BCELoss()
 
         # parameters initialization
@@ -100,17 +102,13 @@ def fwfm_layer(self, infeature):
         """
         # get r(Fi, Fj)
         batch_size = infeature.shape[0]
-        para = torch.randn(self.num_fields * self.num_fields * self.embedding_size).\
-            expand(batch_size, self.num_fields * self.num_fields * self.embedding_size).\
-            to(self.device)  # [batch_size*num_pairs*emb_dim]
-        para = para.reshape(batch_size, self.num_fields, self.num_fields, self.embedding_size)
-        r = nn.Parameter(para, requires_grad=True)  # [batch_size, num_fields, num_fields, emb_dim]
+        weight = self.weight.expand(batch_size,-1,-1,-1)
 
         fwfm_inter = list()  # [batch_size, num_fields, emb_dim]
         for i in range(self.num_features - 1):
             for j in range(i + 1, self.num_features):
                 Fi, Fj = self.feature2field[i], self.feature2field[j]
-                fwfm_inter.append(infeature[:, i] * infeature[:, j] * r[:, Fi, Fj])
+                fwfm_inter.append(infeature[:, i] * infeature[:, j] * weight[:, Fi, Fj])
         fwfm_inter = torch.stack(fwfm_inter, dim=1)
         fwfm_inter = torch.sum(fwfm_inter, dim=1)  # [batch_size, emb_dim]
         fwfm_inter = self.dropout_layer(fwfm_inter)

recbole/model/layers.py

@@ -269,7 +269,7 @@ def __init__(
         self.softmax_stag = softmax_stag
         self.return_seq_weight = return_seq_weight
         self.mask_mat = mask_mat
-        self.att_mlp_layers = MLPLayers(self.att_hidden_size, activation='Sigmoid', bn=False)
+        self.att_mlp_layers = MLPLayers(self.att_hidden_size, activation=self.activation, bn=False)
         self.dense = nn.Linear(self.att_hidden_size[-1], 1)
 
     def forward(self, queries, keys, keys_length):