Skip to content

Commit e307916

Browse files
committed
fix
1 parent 49e669f commit e307916

File tree

2 files changed

+2
-2
lines changed

2 files changed

+2
-2
lines changed

models/recall/ncf/config_bigdata.yaml

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -15,7 +15,7 @@
1515
runner:
1616
train_data_dir: "../../../datasets/movielens_pinterest_NCF/big_train"
1717
train_reader_path: "movielens_reader" # importlib format
18-
train_batch_size: 1024
18+
train_batch_size: 256
1919
model_save_path: "output_model_ncf"
2020

2121
use_gpu: False

models/recall/ncf/readme.md

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -81,7 +81,7 @@ python -u ../../../tools/static_infer.py -m config.yaml
8181

8282
| 模型 | HR@10 | NDCG@10 | batch_size | epoch_num| Time of each epoch |
8383
| :------| :------ |:------ | :------ | :------| :------ |
84-
| NCF_NeuMF | 0.58 | 0.33 | 1024 | 20 | 约20分钟 |
84+
| NCF_NeuMF | 0.58 | 0.33 | 256 | 20 | 约20分钟 |
8585

8686
1. 确认您当前所在目录为PaddleRec/models/recall/ncf
8787
2. 进入paddlerec/datasets/movielens_pinterest_NCF目录下,执行该脚本,会从国内源的服务器上下载我们预处理完成的movielens和pinterest全量数据集,并解压到指定文件夹。

0 commit comments

Comments
 (0)