Merge pull request #688 from wangzhen38/fix_bert4rec

fix_bert4rec_format

Author: frankwhzhang

models/rank/bert4rec/data_augment_candi_gen.py renamed to datasets/Beauty/data_augment_candi_gen.py

@@ -0,0 +1,4 @@
+mkdir data
+wget https://paddlerec.bj.bcebos.com/datasets/Beauty/beauty.txt
+mv beauty.txt ./data/
+python data_augment_candi_gen.py

datasets/Beauty/run.sh

@@ -0,0 +1,53 @@
+# 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.
+
+# global settings
+
+runner:
+  train_data_dir: "data/train/"
+  train_reader_path: "data_reader" # importlib format
+  use_gpu: False
+  train_batch_size: 1
+  data_batch_size: 1
+  epochs: 1
+  print_interval: 1
+
+  model_save_path: "output_model_bert4rec_demo"
+  test_data_dir: "data/test/"
+  infer_reader_path: "data_reader" # importlib format
+  infer_batch_size: 1
+  infer_load_path: "output_model_bert4rec_demo"
+  infer_start_epoch: 0
+  infer_end_epoch: 1
+
+
+# hyper parameters of user-defined network
+hyper_parameters:
+  # optimizer config
+  optimizer:
+    learning_rate: 0.0001
+    weight_decay: 0.01
+
+  _emb_size: 64
+  _n_layer: 2
+  _n_head: 2
+  _voc_size: 54546
+  _max_position_seq_len: 50
+  _sent_types: 2
+  hidden_act: "gelu"
+  _dropout: 0.5
+  _attention_dropout: 0.2
+  _param_initializer: 0.02
+  num_test_user: 1 
+

models/rank/bert4rec/config.yaml

@@ -15,19 +15,19 @@
 # global settings
 
 runner:
-  train_data_dir: "data/train"
+  train_data_dir: "../../../datasets/Beauty/data/train"
   train_reader_path: "data_reader" # importlib format
   use_gpu: True
   train_batch_size: 1
   data_batch_size: 256
   epochs: 10
   print_interval: 100
 
-  model_save_path: "output_model_bert4rec"
-  test_data_dir: "data/test"
+  model_save_path: "output_model_bert4rec_all"
+  test_data_dir: "../../../datasets/Beauty/data/test"
   infer_reader_path: "data_reader" # importlib format
   infer_batch_size: 1
-  infer_load_path: "output_model_bert4rec"
+  infer_load_path: "output_model_bert4rec_all"
   infer_start_epoch: 9
   infer_end_epoch: 10