Merge pull request #782 from wangzhen38/add_escm

Add escm

Author: frankwhzhang

README_CN.md

@@ -179,6 +179,7 @@ python -u tools/static_trainer.py -m models/rank/dnn/config.yaml #  静态图训
   |  多任务  |                     [ShareBottom](models/multitask/share_bottom/)([文档](https://paddlerec.readthedocs.io/en/latest/models/multitask/share_bottom.html))                      |  [Python CPU/GPU](https://aistudio.baidu.com/aistudio/projectdetail/3238943)  |       ✓     |     ✓     | >=2.1.0 | [1998][Multitask learning](http://reports-archive.adm.cs.cmu.edu/anon/1997/CMU-CS-97-203.pdf)                                                                                                               |
   |  多任务  |                                 [Maml](models/multitask/maml/)([文档](https://paddlerec.readthedocs.io/en/latest/models/multitask/maml.html))                                 |  [Python CPU/GPU](https://aistudio.baidu.com/aistudio/projectdetail/3238412)  |      x      |     x     | >=2.1.0 | [PMLR 2017][Model-agnostic meta-learning for fast adaptation of deep networks](https://arxiv.org/pdf/1703.03400.pdf)                                                                                                               |
   |  多任务  |                         [DSelect_K](models/multitask/dselect_k/)([文档](https://paddlerec.readthedocs.io/en/latest/models/multitask/dselect_k.html))                          |  -  |      x      |     x     | >=2.1.0 | [NeurIPS 2021][DSelect-k: Differentiable Selection in the Mixture of Experts with Applications to Multi-Task Learning](https://arxiv.org/pdf/2106.03760v3.pdf)                                                                                                               |
+  |  多任务  |                         [ESCM2](models/multitask/escm2/)                          |  -  |      x      |     x     | >=2.1.0 | [SIGIR 2022][ESCM2: Entire Space Counterfactual Multi-Task Model for Post-Click Conversion Rate Estimation](https://arxiv.org/pdf/2204.05125.pdf)                                                                                                               |
   |  重排序  |                                      [Listwise](https://github.com/PaddlePaddle/PaddleRec/tree/release/1.8.5/models/rerank/listwise/)                                       |  -  |       ✓     |     x     | [1.8.5](https://github.com/PaddlePaddle/PaddleRec/tree/release/1.8.5) | [2019][Sequential Evaluation and Generation Framework for Combinatorial Recommender System](https://arxiv.org/pdf/1902.00245.pdf)                                                                           |
 
 

README_EN.md

@@ -170,6 +170,7 @@ python -u tools/static_trainer.py -m models/rank/dnn/config.yaml #  Training wit
   |      Multi-Task       |           [ShareBottom](models/multitask/share_bottom/)<br>([doc](https://paddlerec.readthedocs.io/en/latest/models/multitask/share_bottom.html))           |  [Python CPU/GPU](https://aistudio.baidu.com/aistudio/projectdetail/3238943)  |         ✓         |     ✓     |  >=2.1.0 | [1998][Multitask learning](http://reports-archive.adm.cs.cmu.edu/anon/1997/CMU-CS-97-203.pdf)                                                                                                               |
   |      Multi-Task       |           [Maml](models/multitask/maml/)<br>([doc](https://paddlerec.readthedocs.io/en/latest/models/multitask/maml.html))           |  [Python CPU/GPU](https://aistudio.baidu.com/aistudio/projectdetail/3238412)  |    x      |     x     | >=2.1.0 | [PMLR 2017][Model-agnostic meta-learning for fast adaptation of deep networks](https://arxiv.org/pdf/1703.03400.pdf)                                                                                                               |
   |  Multi-Task  |           [DSelect_K](models/multitask/dselect_k/)<br>([doc](https://paddlerec.readthedocs.io/en/latest/models/multitask/dselect_k.html))           |  -  |      x      |     x     | >=2.1.0 | [NeurIPS 2021][DSelect-k: Differentiable Selection in the Mixture of Experts with Applications to Multi-Task Learning](https://arxiv.org/pdf/2106.03760v3.pdf)                                                                                                               |
+  |  Multi-Task  |                         [ESCM2](models/multitask/escm2/)                          |  -  |      x      |     x     | >=2.1.0 | [SIGIR 2022][ESCM2: Entire Space Counterfactual Multi-Task Model for Post-Click Conversion Rate Estimation](https://arxiv.org/pdf/2204.05125.pdf)                                                                                                               |
   |        Re-Rank        |                [Listwise](https://github.com/PaddlePaddle/PaddleRec/tree/release/1.8.5/models/rerank/listwise/)                |  -  |         ✓         |     x     |  [1.8.5](https://github.com/PaddlePaddle/PaddleRec/tree/release/1.8.5) | [2019][Sequential Evaluation and Generation Framework for Combinatorial Recommender System](https://arxiv.org/pdf/1902.00245.pdf)                                                                           |
 
 <h2 align="center">Community</h2>

doc/imgs/escm2.png

@@ -0,0 +1,13 @@
+# Copyright (c) 2020 PaddlePaddle Authors. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.

doc/imgs/escm2_dr1.png

@@ -0,0 +1,56 @@
+# Copyright (c) 2020 PaddlePaddle Authors. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+
+runner:
+  train_data_dir: "data/train"
+  train_reader_path: "escm_reader" # importlib format
+  use_gpu: False
+  use_auc: True
+  auc_num: 2
+  train_batch_size: 2
+  epochs: 3
+  print_interval: 2
+  #model_init_path: "output_model_escm/2" # init model
+  model_save_path: "output_model_escm"
+  test_data_dir: "data/train"
+  infer_batch_size: 2
+  infer_reader_path: "escm_reader" # importlib format
+  infer_load_path: "output_model_escm"
+  infer_start_epoch: 0
+  infer_end_epoch: 3
+  counterfact_mode: "DR"
+  #use inference save model
+  use_inference: False
+  save_inference_feed_varnames: ["field_0", "field_1", "field_2", "field_3", "field_4", "field_5", "field_6", "field_7", "field_8", "field_9", "field_10", "field_11", "field_12", "field_13", "field_14", "field_15", "field_16", "field_17", "field_18", "field_19", "field_20", "field_21", "field_22"]
+  save_inference_fetch_varnames: ["softmax_0.tmp_0", "concat_1.tmp_0"]
+
+hyper_parameters:
+  sparse_feature_number: 737946
+  sparse_feature_dim: 12
+  num_field: 23
+  ctr_fc_sizes: [256, 64]
+  cvr_fc_sizes: [256, 64]
+  global_w: 1.0
+  counterfactual_w: 0.01
+  feature_size: 276 
+  expert_num: 8
+  gate_num: 3
+  expert_size: 16
+  tower_size: 8
+
+  optimizer:
+    class: adam
+    learning_rate: 0.001
+    strategy: async

doc/imgs/escm2_dr2.png

@@ -0,0 +1,57 @@
+# Copyright (c) 2020 PaddlePaddle Authors. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+
+runner:
+  train_data_dir: "../../../datasets/ali-ccp/train_data"
+  train_reader_path: "escm_reader" # importlib format
+  use_gpu: True
+  use_auc: True
+  auc_num: 2
+  train_batch_size: 1024
+  epochs: 10
+  print_interval: 10
+  #model_init_path: "output_model/0" # init model
+  model_save_path: "output_model_escm_all"
+  test_data_dir: "../../../datasets/ali-ccp/test_data"
+  infer_batch_size: 1024
+  infer_reader_path: "escm_reader" # importlib format
+  infer_load_path: "output_model_escm_all"
+  infer_start_epoch: 0
+  infer_end_epoch: 10
+  counterfact_mode: "DR"
+  #use inference save model
+  use_inference: False
+  save_inference_feed_varnames: ["field_0", "field_1", "field_2", "field_3", "field_4", "field_5", "field_6", "field_7", "field_8", "field_9", "field_10", "field_11", "field_12", "field_13", "field_14", "field_15", "field_16", "field_17", "field_18", "field_19", "field_20", "field_21", "field_22"]
+  save_inference_fetch_varnames: ["softmax_0.tmp_0", "concat_1.tmp_0"]
+
+
+hyper_parameters:
+  sparse_feature_number: 737946
+  sparse_feature_dim: 12
+  num_field: 23
+  ctr_fc_sizes: [256, 64]
+  cvr_fc_sizes: [256, 64]
+  global_w: 0.5
+  counterfactual_w: 0.5
+  expert_num: 8
+  gate_num: 2
+  expert_size: 16
+  tower_size: 8
+  feature_size: 276
+
+  optimizer:
+    class: adam
+    learning_rate: 0.001
+    strategy: async