Dev swc (shenweichen#377)

- enhance the compatibility for tf 2.5
- add training param to activation layer
- add PositionEncoding Layer

Author: shenweichen

.github/ISSUE_TEMPLATE/bug_report.md

@@ -8,7 +8,7 @@ assignees: ''
 ---
 
 **Describe the bug(问题描述)**
-A clear and concise description of what the bug is.
+A clear and concise description of what the bug is.Better with standalone code to reproduce the issue.
 
 **To Reproduce(复现步骤)**
 Steps to reproduce the behavior:
@@ -18,9 +18,9 @@ Steps to reproduce the behavior:
 4. See error
 
 **Operating environment(运行环境):**
- - python version [e.g. 3.5, 3.7]
- - tensorflow version [e.g. 1.4.0, 1.15.0, 2.4.0]
- - deepctr version [e.g. 0.8.3,]
+ - python version [e.g. 3.6, 3.7]
+ - tensorflow version [e.g. 1.4.0, 1.15.0, 2.5.0]
+ - deepctr version [e.g. 0.8.6,]
 
 **Additional context**
 Add any other context about the problem here.

.github/ISSUE_TEMPLATE/question.md

@@ -16,5 +16,5 @@ Add any other context about the problem here.
 
 **Operating environment(运行环境):**
  - python version [e.g. 3.6]
- - tensorflow version [e.g. 1.4.0, 1.5.0, 2.4.0]
- - deepctr version [e.g. 0.8.3,]
+ - tensorflow version [e.g. 1.4.0, 1.15.0, 2.5.0]
+ - deepctr version [e.g. 0.8.6,]

.github/workflows/ci.yml

@@ -14,15 +14,17 @@ jobs:
   build:
 
     runs-on: ubuntu-latest
-    timeout-minutes: 120
+    timeout-minutes: 180
     strategy:
       matrix:
         python-version: [3.6,3.7]
-        tf-version: [1.4.0,1.15.0,2.1.0,2.4.0]
+        tf-version: [1.4.0,1.15.0,2.1.0,2.5.0]
 
         exclude:
           - python-version: 3.7
             tf-version: 1.4.0
+          - python-version: 3.7
+            tf-version: 1.15.0
 
     steps:
 
@@ -40,7 +42,7 @@ jobs:
         pip install -q requests
         pip install -e .
     - name: Test with pytest
-      timeout-minutes: 120
+      timeout-minutes: 180
       run: |
         pip install -q pytest
         pip install -q pytest-cov

README.md

@@ -11,7 +11,7 @@
 
 [![Documentation Status](https://readthedocs.org/projects/deepctr-doc/badge/?version=latest)](https://deepctr-doc.readthedocs.io/)
 ![CI status](https://github.com/shenweichen/deepctr/workflows/CI/badge.svg)
-[![Coverage Status](https://coveralls.io/repos/github/shenweichen/DeepCTR/badge.svg?branch=master)](https://coveralls.io/github/shenweichen/DeepCTR?branch=master)
+[![codecov](https://codecov.io/gh/shenweichen/DeepCTR/branch/master/graph/badge.svg)](https://codecov.io/gh/shenweichen/DeepCTR)
 [![Codacy Badge](https://api.codacy.com/project/badge/Grade/d4099734dc0e4bab91d332ead8c0bdd0)](https://www.codacy.com/gh/shenweichen/DeepCTR?utm_source=github.com&utm_medium=referral&utm_content=shenweichen/DeepCTR&utm_campaign=Badge_Grade)
 [![Disscussion](https://img.shields.io/badge/chat-wechat-brightgreen?style=flat)](./README.md#DisscussionGroup)
 [![License](https://img.shields.io/github/license/shenweichen/deepctr.svg)](https://github.com/shenweichen/deepctr/blob/master/LICENSE)
@@ -81,7 +81,7 @@ If you find this code useful in your research, please cite it using the followin
 ## DisscussionGroup
 
 - [Discussions](https://github.com/shenweichen/DeepCTR/discussions)
-- 公众号：**浅梦的学习笔记**  
+- 公众号：**浅梦学习笔记**  
 - wechat ID: **deepctrbot**
 
   ![wechat](./docs/pics/code.png)
@@ -115,7 +115,7 @@ If you find this code useful in your research, please cite it using the followin
       </td>
       <td>
         ​ <a href="https://github.com/TanTingyi"><img width="70" height="70" src="https://github.com/TanTingyi.png?s=40" alt="pic"></a><br>
-         <a href="https://github.com/TanTingyi">LeoCai</a>
+         <a href="https://github.com/TanTingyi">Tan Tingyi</a>
          <p>  Chongqing University <br> of  Posts and <br> Telecommunications   </p>​
       </td>
     </tr>

deepctr/__init__.py

@@ -1,4 +1,4 @@
 from .utils import check_version
 
-__version__ = '0.8.5'
+__version__ = '0.8.6'
 check_version(__version__)

deepctr/estimator/models/deepfefm.py

@@ -18,17 +18,16 @@
 from ...layers.utils import concat_func, add_func, combined_dnn_input, reduce_sum
 
 
-def DeepFEFMEstimator(linear_feature_columns, dnn_feature_columns, embedding_size=48,
-                      dnn_hidden_units=(1024, 1024, 1024), l2_reg_linear=0.000001, l2_reg_embedding_feat=0.00001,
-                      l2_reg_embedding_field=0.0000001, l2_reg_dnn=0, seed=1024, dnn_dropout=0.2,
+def DeepFEFMEstimator(linear_feature_columns, dnn_feature_columns,
+                      dnn_hidden_units=(128, 128), l2_reg_linear=0.00001, l2_reg_embedding_feat=0.00001,
+                      l2_reg_embedding_field=0.00001, l2_reg_dnn=0, seed=1024, dnn_dropout=0.0,
                       dnn_activation='relu', dnn_use_bn=False, task='binary', model_dir=None,
                       config=None, linear_optimizer='Ftrl', dnn_optimizer='Adagrad', training_chief_hooks=None):
     """Instantiates the DeepFEFM Network architecture or the shallow FEFM architecture (Ablation support not provided
     as estimator is meant for production, Ablation support provided in DeepFEFM implementation in models
 
     :param linear_feature_columns: An iterable containing all the features used by linear part of the model.
     :param dnn_feature_columns: An iterable containing all the features used by deep part of the model.
-    :param embedding_size: positive integer,sparse feature embedding_size
     :param dnn_hidden_units: list,list of positive integer or empty list, the layer number and units in each layer of DNN
     :param l2_reg_linear: float. L2 regularizer strength applied to linear part
     :param l2_reg_embedding_feat: float. L2 regularizer strength applied to embedding vector of features
@@ -62,10 +61,11 @@ def _model_fn(features, labels, mode, config):
             sparse_embedding_list, dense_value_list = input_from_feature_columns(features, dnn_feature_columns,
                                                                                  l2_reg_embedding=l2_reg_embedding_feat)
 
-            fefm_interaction_embedding = FEFMLayer(num_fields=len(sparse_embedding_list), embedding_size=embedding_size,
-                                   regularizer=l2_reg_embedding_field)(concat_func(sparse_embedding_list, axis=1))
+            fefm_interaction_embedding = FEFMLayer(
+                regularizer=l2_reg_embedding_field)(concat_func(sparse_embedding_list, axis=1))
 
-            fefm_logit = tf.keras.layers.Lambda(lambda x: reduce_sum(x, axis=1, keep_dims=True))(fefm_interaction_embedding)
+            fefm_logit = tf.keras.layers.Lambda(lambda x: reduce_sum(x, axis=1, keep_dims=True))(
+                fefm_interaction_embedding)
 
             final_logit_components.append(fefm_logit)
 
@@ -87,6 +87,3 @@ def _model_fn(features, labels, mode, config):
                                 training_chief_hooks=training_chief_hooks)
 
     return tf.estimator.Estimator(_model_fn, model_dir=model_dir, config=config)
-
-
-

deepctr/estimator/utils.py

@@ -44,27 +44,29 @@ def _eval_metric_ops(self,
             _summary_key(self._name, "prediction/mean"): metrics.mean(predictions, weights=weights),
             _summary_key(self._name, "label/mean"): metrics.mean(labels, weights=weights),
         }
-        tf.summary.scalar("prediction/mean", metric_ops[_summary_key(self._name, "prediction/mean")][1])
-        tf.summary.scalar("label/mean", metric_ops[_summary_key(self._name, "label/mean")][1])
+
+        summary_scalar("prediction/mean", metric_ops[_summary_key(self._name, "prediction/mean")][1])
+        summary_scalar("label/mean", metric_ops[_summary_key(self._name, "label/mean")][1])
+
 
         mean_loss = losses.compute_weighted_loss(
             unweighted_loss, weights=1.0, reduction=losses.Reduction.MEAN)
 
         if self._task == "binary":
             metric_ops[_summary_key(self._name, "LogLoss")] = metrics.mean(mean_loss, weights=weights, )
-            tf.summary.scalar("LogLoss", mean_loss)
+            summary_scalar("LogLoss", mean_loss)
 
             metric_ops[_summary_key(self._name, "AUC")] = metrics.auc(labels, predictions, weights=weights)
-            tf.summary.scalar("AUC", metric_ops[_summary_key(self._name, "AUC")][1])
+            summary_scalar("AUC", metric_ops[_summary_key(self._name, "AUC")][1])
         else:
 
             metric_ops[_summary_key(self._name, "MSE")] = metrics.mean_squared_error(labels, predictions,
                                                                                      weights=weights)
-            tf.summary.scalar("MSE", mean_loss)
+            summary_scalar("MSE", mean_loss)
 
             metric_ops[_summary_key(self._name, "MAE")] = metrics.mean_absolute_error(labels, predictions,
                                                                                       weights=weights)
-            tf.summary.scalar("MAE", metric_ops[_summary_key(self._name, "MAE")][1])
+            summary_scalar("MAE", metric_ops[_summary_key(self._name, "MAE")][1])
 
         return metric_ops
 
@@ -206,3 +208,10 @@ def to_float(x, name="ToFloat"):
         return tf.to_float(x, name)
     except AttributeError:
         return tf.compat.v1.to_float(x, name)
+
+
+def summary_scalar(name, data):
+    try:
+        tf.summary.scalar(name, data)
+    except AttributeError:  # tf version 2.5.0+:AttributeError: module 'tensorflow._api.v2.summary' has no attribute 'scalar'
+        tf.compat.v1.summary.scalar(name, data)

deepctr/feature_column.py

@@ -9,7 +9,7 @@
 from .inputs import create_embedding_matrix, embedding_lookup, get_dense_input, varlen_embedding_lookup, \
     get_varlen_pooling_list, mergeDict
 from .layers import Linear
-from .layers.utils import concat_func, add_func
+from .layers.utils import concat_func
 
 DEFAULT_GROUP_NAME = "default_group"
 

deepctr/layers/__init__.py

@@ -9,7 +9,7 @@
 from .normalization import LayerNormalization
 from .sequence import (AttentionSequencePoolingLayer, BiasEncoding, BiLSTM,
                        KMaxPooling, SequencePoolingLayer, WeightedSequenceLayer,
-                       Transformer, DynamicGRU)
+                       Transformer, DynamicGRU,PositionEncoding)
 
 from .utils import NoMask, Hash, Linear, Add, combined_dnn_input, softmax, reduce_sum
 
@@ -47,5 +47,6 @@
                   'FwFMLayer': FwFMLayer,
                   'softmax': softmax,
                   'FEFMLayer': FEFMLayer,
-                  'reduce_sum': reduce_sum
+                  'reduce_sum': reduce_sum,
+                  'PositionEncoding':PositionEncoding
                   }

deepctr/layers/core.py

@@ -189,8 +189,11 @@ def call(self, inputs, training=None, **kwargs):
 
             if self.use_bn:
                 fc = self.bn_layers[i](fc, training=training)
-
-            fc = self.activation_layers[i](fc)
+            try:
+                fc = self.activation_layers[i](fc, training=training)
+            except TypeError as e:  # TypeError: call() got an unexpected keyword argument 'training'
+                print("make sure the activation function use training flag properly", e)
+                fc = self.activation_layers[i](fc)
 
             fc = self.dropout_layers[i](fc, training=training)
             deep_input = fc