- Fix TransAct bug (#132)

- Set polars <= 1.0.0
- Remove group_id from feature_map.json

README.md

@@ -81,8 +81,8 @@ Click-through rate (CTR) prediction is a critical task for various industrial ap
 | 44  | DLP-KDD'19        | [BST](./model_zoo/BST)                   | [Behavior Sequence Transformer for E-commerce Recommendation in Alibaba](https://arxiv.org/abs/1905.06874) :triangular_flag_on_post:**Alibaba**                                                                 |  [:arrow_upper_right:](https://github.com/reczoo/BARS/tree/main/ranking/ctr/BST)     | `torch`       |
 | 45  | CIKM'20           | [DMIN](./model_zoo/DMIN)                 | [Deep Multi-Interest Network for Click-through Rate Prediction](https://dl.acm.org/doi/10.1145/3340531.3412092) :triangular_flag_on_post:**Alibaba**                                                            | [:arrow_upper_right:](https://github.com/reczoo/BARS/tree/main/ranking/ctr/DMIN)                                                                                                                 | `torch`       |
 | 46  | AAAI'20           | [DMR](./model_zoo/DMR)                   | [Deep Match to Rank Model for Personalized Click-Through Rate Prediction](https://ojs.aaai.org/index.php/AAAI/article/view/5346) :triangular_flag_on_post:**Alibaba**                                           |    [:arrow_upper_right:](https://github.com/reczoo/BARS/tree/main/ranking/ctr/DMR)                                                                                                                  | `torch`       |
-| 47  | DLP-KDD'22          | [ETA](./model_zoo/ETA)                   | [Efficient Long Sequential User Data Modeling for Click-Through Rate Prediction](https://arxiv.org/abs/2209.12212) :triangular_flag_on_post:**Alibaba**                                                               |                                                                                                                 | `torch`       |
-| 48  | CIKM'22           | [SDIM](./model_zoo/SDIM)                 | [Sampling Is All You Need on Modeling Long-Term User Behaviors for CTR Prediction](https://arxiv.org/abs/2205.10249) :triangular_flag_on_post:**Meituan**                                                       |                                                                                                                 | `torch`       |
+| 47  | DLP-KDD'22          | [ETA](./model_zoo/LongCTR/ETA)                   | [Efficient Long Sequential User Data Modeling for Click-Through Rate Prediction](https://arxiv.org/abs/2209.12212) :triangular_flag_on_post:**Alibaba**                                                               |                                                                                                                 | `torch`       |
+| 48  | CIKM'22           | [SDIM](./model_zoo/LongCTR/SDIM)                 | [Sampling Is All You Need on Modeling Long-Term User Behaviors for CTR Prediction](https://arxiv.org/abs/2205.10249) :triangular_flag_on_post:**Meituan**                                                       |                                                                                                                 | `torch`       |
 | 49  | KDD'23           | [TransAct](./model_zoo/TransAct)                 | [TransAct: Transformer-based Realtime User Action Model for Recommendation at Pinterest](https://arxiv.org/abs/2306.00248) :triangular_flag_on_post:**Pinterest**                                                       | [:arrow_upper_right:](https://github.com/reczoo/BARS/tree/main/ranking/ctr/TransAct)         | `torch`       |
 |<tr><th colspan=6 align="center">:open_file_folder: **Dynamic Weight Network**</th></tr>|
 | 50  | NeurIPS'22          | [APG](./model_zoo/APG)               | [APG: Adaptive Parameter Generation Network for Click-Through Rate Prediction](https://arxiv.org/abs/2203.16218) :triangular_flag_on_post:**Alibaba**                                |    [:arrow_upper_right:](https://github.com/reczoo/BARS/tree/main/ranking/ctr/APG)                                                                                                   | `torch`       |

fuxictr/features.py

@@ -45,7 +45,7 @@ def load(self, json_file, params):
         self.labels = feature_map.get("labels", [])
         self.total_features = feature_map.get("total_features", 0)
         self.input_length = feature_map.get("input_length", 0)
-        self.group_id = feature_map.get("group_id", None)
+        self.group_id = params.get("group_id", None)
         self.default_emb_dim = params.get("embedding_dim", None)
         self.features = OrderedDict((k, v) for x in feature_map["features"] for k, v in x.items())
         self.num_fields = self.get_num_fields()
@@ -74,8 +74,6 @@ def save(self, json_file):
         feature_map["total_features"] = self.total_features
         feature_map["input_length"] = self.input_length
         feature_map["labels"] = self.labels
-        if self.group_id is not None:
-            feature_map["group_id"] = self.group_id
         feature_map["features"] = [{k: v} for k, v in self.features.items()]
         with open(json_file, "w") as fd:
             json.dump(feature_map, fd, indent=4)

model_zoo/LongCTR/ETA/config/dataset_config.yaml

@@ -1,8 +1,8 @@
 ### Tiny data for tests only
 tiny_seq:
-    data_root: ../../data/
+    data_root: ../../../data/
     data_format: npz
-    train_data: ../../data/tiny_seq/train.npz
-    valid_data: ../../data/tiny_seq/valid.npz
-    test_data: ../../data/tiny_seq/test.npz
+    train_data: ../../../data/tiny_seq/train.npz
+    valid_data: ../../../data/tiny_seq/valid.npz
+    test_data: ../../../data/tiny_seq/test.npz
 

model_zoo/LongCTR/SDIM/config/dataset_config.yaml

@@ -1,8 +1,8 @@
 ### Tiny data for tests only
 tiny_seq:
-    data_root: ../../data/
+    data_root: ../../../data/
     data_format: npz
-    train_data: ../../data/tiny_seq/train.npz
-    valid_data: ../../data/tiny_seq/valid.npz
-    test_data: ../../data/tiny_seq/test.npz
+    train_data: ../../../data/tiny_seq/train.npz
+    valid_data: ../../../data/tiny_seq/valid.npz
+    test_data: ../../../data/tiny_seq/test.npz
 

model_zoo/TransAct/src/TransAct.py

@@ -122,7 +122,6 @@ def __init__(self,
                 else embedding_dim
             )
             transformer_in_dim = seq_emb_dim + target_emb_dim
-            seq_out_dim += (first_k_cols + int(concat_max_pool)) * transformer_in_dim
             self.transformer_encoders.append(
                 TransActTransformer(transformer_in_dim,
                                     dim_feedforward=dim_feedforward,
@@ -134,7 +133,8 @@ def __init__(self,
                                     first_k_cols=first_k_cols,
                                     concat_max_pool=concat_max_pool)
             )
-        dcn_in_dim = feature_map.sum_emb_out_dim() + seq_out_dim - seq_emb_dim
+            seq_out_dim += (first_k_cols + int(concat_max_pool)) * transformer_in_dim - seq_emb_dim
+        dcn_in_dim = feature_map.sum_emb_out_dim() + seq_out_dim
         self.crossnet = CrossNetV2(dcn_in_dim, dcn_cross_layers)
         self.parallel_dnn = MLP_Block(input_dim=dcn_in_dim,
                                       output_dim=None, # output hidden layer

requirements.txt

@@ -6,4 +6,4 @@ numpy
 h5py
 tqdm
 pyarrow
-polars
+polars<=1.0.0